Archaeology Artifacts and Antiquities of The Ancient Near East PDF

Archaeology, Artifacts and
Antiquities of the Ancient Near East

Culture and History of the
Ancient Near East
Founding Editor
M.H.E. Weippert
Editor-in-Chief
Thomas Schneider
Editors
Eckart Frahm (Yale University)
W. Randall Garr (University of California, Santa Barbara)
B. Halpern (Pennsylvania State University)
Theo P.J. van den Hout (Oriental Institute)
Irene J. Winter (Harvard University)
VOLUME 62
The titles published in this series are listed at brill.com/chan

Archaeology, Artifacts and
Antiquities of the Ancient Near East
Sites, Cultures, and Proveniences
By
Oscar White Muscarella
LEIDEN BOSTON
2013
Library of Congress Cataloging-in-Publication Data
Muscarella, Oscar White.

Archaeology, artifacts and antiquities of the ancient Near East : sites, cultures, and proveniences /
by Oscar White Muscarella.
p. cm.
ISBN 978-90-04-23666-0 (hardback : alk. paper) ISBN 978-90-04-23669-1 (e-book) 1. Middle
EastAntiquities. 2. Excavations (Archaeology)Middle East. 3. Middle EastCivilizationTo 622. 4.
IranAntiquities. 5. Excavations (Archaeology)Iran. 6. TurkeyAntiquities. 7. Excavations
(Archaeology)TurkeyGordion (Extinct city) 8. Forgery of antiquities. I. Title.
DS56.M848 2013
939.4dc23
2013006440
This publication has been typeset in the multilingual Brill typeface. With over 5,100
characters covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for
use in the humanities. For more information, please see www.brill.com/brill-typeface.
ISSN 1566-2055
ISBN 978-90-04-23666-0 (hardback)
ISBN 978-90-04-23669-1 (e-book)
Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing,
IDC Publishers and Martinus Nijhoff Publishers.
All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV
provided that the appropriate fees are paid directly to The Copyright Clearance Center,
222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.
Fees are subject to change.
This book is printed on acid-free paper.

CONTENTS
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PART ONE
SITES AND EXCAVATIONS
Section One. Iran

1. The Tumuli at S Girdan: A Preliminary Report . . . . . . . . . . . . . . . . . . . . 9
2. The Tumuli at S Girdan: Second Report . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3. The Chronology and Culture of S Girdan: Phase III . . . . . . . . . . . . . . . 83
4. Qalatgah: An Urartian Site in Northwestern Iran . . . . . . . . . . . . . . . . . . . 97
5. Excavations at Agrab Tepe, Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6. The Iron Age at Dinkha Tepe, Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7. Warfare at Hasanlu in the Late 9th Century bc . . . . . . . . . . . . . . . . . . . . . 255
8. The Hasanlu Lion Pins Again . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
9. The Excavation of Hasanlu: An Archaeological Evaluation . . . . . . . . . 305
10. The Iranian Iron III Chronology at Muweilah in the Emirate of
Sharjah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11. The Location of Ulhu and Uise in Sargon IIs Eighth Campaign,
714bc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
12. Surkh Dum at The Metropolitan Museum of Art: A Mini-Report . . . 389
13. North-Western Iran: Bronze Age to Iron Age . . . . . . . . . . . . . . . . . . . . . . . 459
14. Jiroft and Jiroft-Aratta: A Review Article of Yousef Madjidzadeh,
Jiroft: The Earliest Oriental Civilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
15. Sargon IIs 8th Campaign: An Introduction and Overview. . . . . . . . . . 523
Section Two. Anatolia

16. King Midas Tumulus at Gordion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
17. The Iron Age Background to the Formation of the Phrygian State . . 549
18. The Date of the Destruction of the Early Phrygian Period at
Gordion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
19. Again Gordions Early Phrygian Destruction Date: ca. 700 +/- bc . . . 601
20. Urartian Metal Artifacts: An Archaeological Review . . . . . . . . . . . . . . . 621
vi contents
PART TWO
ARTIFACTS, CULTURES, FORGERIES, AND PROVENIENCE
Section One. The Aegean and the Ancient Near East

21. The Archaeological Evidence for Relations between Greece and
Iran in the First Millennium bc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
22. Urartian Bells and Samos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
23. King Midas of Phrygia and the Greeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
24. Greek and Oriental Cauldron Attachments: A Review . . . . . . . . . . . . . 725
Section Two. Artifacts

25. Fibulae Represented on Sculpture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
26. Phrygian or Lydian? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
27. Fibulae and Chronology, Marlik and Assur . . . . . . . . . . . . . . . . . . . . . . . . . 803
28. Parasols in the Ancient Near East . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
Section Three. The Antiquities Market and the Plunder Culture

29. The Pope and the Bitter Fanatic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
30. The Antiquities Trade and the Destruction of Ancient Near
Eastern Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837
31. The Fifth Column within the Archaeological Realm: The Great
Divide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861
Section Four. Forgeries

32. Bazaar Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
33. Excavated in the Bazaar: Ashurbanipals Beaker . . . . . . . . . . . . . . . . . . . . 889
34. Von Bissings Memphis Stela: A Product of Cultural Transfer? . . . . . . 901
35. Gudea or not Gudea in New York and Detroit: Ancient or
Modern? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
36. The Veracity of Scientific Testing by Conservators . . . . . . . . . . . . . . . . 931
Section Five. Forgeries of Archaeological Provenience

37. Ziwiye and Ziwiye: The Forgery of a Provenience . . . . . . . . . . . . . . . . . 955
38. Median Art and Medizing Scholarship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 999
39. Museum Constructions of the Oxus Treasures: Forgeries of
Provenience and Ancient Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1025
40. Excavated and Unexcavated Achaemenian Art . . . . . . . . . . . . . . . . . . . . .1043
INTRODUCTION
When asked by Jennifer Pavelko of Brill USA to compile a volume of my

past articles, on the model of the two volumes of Irene Winters outstand-
ing scholarship published by Brill, I was of course flattered. And then began
the difficult task of decision making. Given the limitation of approximately
forty articles to be selected by me, I had first to examine my curriculum
vita and make a list of which to choose. When I reached fifty articles I was
obliged to refocus my mind and eliminate ten of them. This exercise was
fortunate, I believe, for it forced me not merely to focus on works that to my
mind reflect a diachronic compendium of my decades of writing, but con-
comitantly, and as objectively as possible, to contemplate the very nature
and thrusts of my work as a whole. The scope of my research and writing
has widened from the pure archaeological work of excavation reports and
artifact analyses to include a sharpened interest in the vast number of unex-
cavated antiquitiesand, as I came to realize, of forgeriesas well as asso-
ciated cultural phenomena that often run contrary to archaeological goals.
In this introduction I present my academic and intellectual background as a
way of explaining how my interests multiplied and developed. I vigorously
insist that although to some the subjects of artifact analysis, forgery, and
provenance might appear to be diverse, they are to the contrary not discrete
subsets of archaeological discourse; they collectively form equal components
of the core of archaeological research and conclusion-formation.
I readily identify myself as an archaeologist, one concerned with the
material and historical cultures of the ancient Near East, primarily those of
ancient Iran and Anatolia; I do not further identify myself as a specialist,
such as an anthropologist or art historian, inasmuch as I see their activities
as inherent in archaeological research. Over time I came to the conclusion
that what some call art history is in fact the analysis and cultural evaluation
of artifacts, i.e., normal archaeological activity. I discovered quite soon (in
the 1960s) that many archaeologists I encountered had little knowledge of
artifacts in general (except, of course, pottery) and could not identify them
or understand their manifest cultural and chronological value. At Hasanlu
when studying what we excavated I was teased by being called object-
oriented. These interactions only reinforced my opinion of the importance
of artifact analysis in the work of archaeology, as my articles and books make
clear.
2 introduction
My career as an archaeologist began with my interest in Egyptian archae-

ology, determined by my reading of Breasted and Petrie while I was a student
at Stuyvesant High School in New York City.1 I then attended the City Col-
lege of New York, majoring in history (primarily modern history) because
no archaeology courses were in the curriculum.2 While at CCNY, during the
summer I excavated at two sites in the United States, the Pueblo site at Mesa
Verde in Colorado (1953) and a Mandan site, Swan Creek in South Dakota
(1955). At Swan Creek I received notice that I had been awarded a schol-
arship (including a stipend of $57.50 a month, without which I could not
have accepted the scholarship) in classical archaeology at the University of
Pennsylvania. I had first applied to study Egyptian archaeology but I was
rejected for this endeavor; when asked to choose another course, I replied,
Archaeology. I, who then had no Greek, little German, and no academic
archaeological background, was accepted by Rodney S. Young as his stu-
dent. He never told me his reasons for accepting me but it was a miraculous
gift, for my life was utterly altered. Young was one of the most extraordi-
nary and intelligent archaeologists (and human beings) I have ever encoun-
tered.
In 1957 Young took me and (generously, indeed!) my wife Grace, whom I
had recently married, to Gordion for five months.3 Here I experienced two
significant emotional and intellectual events: living and working with won-
derful, friendly Turkish villagers in Yassihyk, and the months-long exca-
vation of Tumulus MM, a burial enclosure filled with sumptuous artifacts,
built and furnished by King Midas for his father Gordias. But there was more,
for it was here that I met my first fibula, in fact a group of them fastening
the clothing of King Gordias. It was a cerebral discovery that never left me,
for those artifacts, along with the socketed arrowheads I later encountered,
served me and others over the years as a critical Leitmotiv and played a key
role in determining post quem chronologies. Gordion remained my archae-
ological home, so to speak, and it was here that I returned in the summer of
1959, leaving the American School of Classical Archaeology in Athens, where
I was spending a year as a Fulbright scholar. I was not allowed to dig in the
1 The works I read included James Henry Breasted, A History of Egypt (2nd ed.; New York:
Charles Scribner, 1924), and W.M. Flinders Petrie, The Religion of Ancient Egypt (London:
A. Constable & Co., 1906), among others.
2 I attended the evening session because I had to earn my living in the daytime.
3 Grace eventually worked with Piet de Jong and became his assistant as well as a fellow
archaeological artist.
introduction 3
Agora because of unbearable class snobbery and went east. Given the dif-
ferences in these two encounters, it is no wonder that I chose to remain in
the world of ancient Near Eastern archaeology instead of classical.
My first engagement with Iranian archaeology came in 1960, the year I
finished my coursework in classical archaeology with Young and was hired
to teach history at CCNY, beginning in the fall. Solely because of my Gordion
background (I had taken only one course in ancient Near Eastern archaeol-
ogy at Penn), Robert H. Dyson Jr. invited me to come to Hasanlu, in north-
western Iran, for which deed I am indebted to him. Thereafter I excavated at
Hasanlu and eight other Iranian and Anatolian sites, and through this work
I became an ancient Near Eastern archaeologist. My doctoral dissertation
(finalized in 1965 and published in 1967) was of course on Phrygian fibulae,
my first artifactual love. In 1964 I left CCNY for a position in the Department
of Ancient Near Eastern Art at the Metropolitan Museum of Art (MMA),
which I left upon my retirement in 2009.
In my early years at the MMA I continued to excavate and to publish
articles derived from those activities. But after a few years of working con-
tinuously with purchased antiquities in the museum, I began to realize
that something was utterly wrong regarding the nature of how scholars
in the discipline, both at the museum and in academia, published antiq-
uities. The very first time I empirically learned about plundering and its
attendant destruction of ancient sites was in the early 1970s, when I heard
lectures by the Harvard Mesoamerican scholar Clemency Coggins and the
Iranian archaeologist Ezat Negahban on this major but for me hitherto fully
neglected issue. I first then noticed that not only my museum colleagues but
many others failed to make the obvious and necessary distinction between
an excavated artifact and a purchased antiquity acquired from collectors,
antiquity dealers, or the auction houses ubiquitous in Europe, the United
States, and in modern Near Eastern countries. I also learned that this activ-
ity was in fact a worldwide cultural problem. My mind stumbled but I soon
came into contact with a few other similarly minded colleagues, mainly
James Wiseman of Boston University and Ross Holloway of Brown Univer-
sity. In 1973 we formed a new archaeological group, the Association of Field
Archaeologists, and soon thereafter we began to publish The Journal of Field
Archaeology (mashallah, still in existence) through Boston University. The
battle had begun, and it has not ceased. My first article on plunder and
the role of museum trustees, officers, and curators as well as private collec-
tors in the destruction of this planets history appeared in the JFA (1/2, 1974:
221222), and the publications have continued nonstop; my most recent
appeared in 2012.
4 introduction
Following upon this discovery, I also came to realize that a very large
number of antiquities in museums, collections, and manifold publications
worldwide were in fact modern forgeries, commonly baptized and accepted
as ancient artifacts by (too) many scholars and introduced into their discus-
sions, reviews, and analyses of ancient Near Eastern cultures. Alas, I did not
learn of this profound flaw in the archaeological discipline at large (i.e., not
only in ancient Near Eastern studies) from my educational and excavation
background, or from my MMA work. I never heard the word forgery men-
tioned in graduate school or in the field, or by colleagues. What I learned of
these issues derived from my normal research and my hands-on archaeolog-
ical background. I did discover that in the late nineteenth and early decades
of the twentieth century a few scholars had addressed these concerns, but
then the subject ceased to be discussed until recent times.
The volume presented here follows the evolution of my scholarly work
and interests and is divided into several categories, which, as noted above, I
see as interrelated fields and not isolated subsets. The first part deals primar-
ily with excavations and associated artifacts, issues in ancient geography and
the identification of ancient sites in northwest Iran, my research involving
the culture and chronology of the Phrygian capital at Gordion in Anatolia,
and the chronology and Iranian cultural relations of a site in the Emirate
of Sharjah. Part two is wide-ranging, as its title makes evident. Represented
are my interests in Aegean and ancient Near Eastern cultural and political
interconnections, the role of fibulae in revealing cultural and chronological
matters, and the gender-determined usage of parasols and their recogni-
tion in excavated contexts. There follow articles specifically concerned with
what I call the Plunder Culture and the forgery of both objects and their
alleged proveniences. In all cases, bibliographies will, inshallah, lead to fur-
ther investigation by scholars and (to me more important) by students.
PART ONE
SITES AND EXCAVATIONS

Section One
Iran
chapter one
THE TUMULI AT S GIRDAN: A PRELIMINARY REPORT*
On August 25, 1936, Sir Aurel Stein completed a six-day excavation on the
mound called Dinkha Tepe, situated in the Ushnu valley in northwestern
Iran, and moved his camp about three miles to the east across the Gadar
River to a location near the modern village of Cheshm Gl. There he
examined a curious succession of conical mounds stretching in a straight
line at short intervals, known as Seh Gird. Stein first thought that the
mounds might be a series of burial tumuli. After examining them and
finding them to be composed of a hard gravel, the same type of soil found in
the adjacent area, he concluded that they were natural.1 The mounds were,
therefore, not excavated, and Stein moved his camp to another area.2
In the summer of 1966, the writer and Robert H. Dyson, Jr., accompanied
by several members of the Hasanlu Project excavating at Dinkha Tepe,
visited the mounds and concluded that they were in fact tumuli and not
natural formations. At that time nine tumuli were counted; in 1968 a total of
eleven were recorded.3
The tumuli lie about five kilometers east of Dinkha Tepe and may be seen
with the naked eye from that mound. They were built about one kilometer
below and west of the foothills that form the eastern boundary of the valley.
The most important site recognized in the immediate area is a recently
discovered Urartian city located at a place in the same foothills now called
* Excerpted from Metropolitan Museum of Art Journal. Copyright 1969 by The Metro-
politan Museum of Art, New York. Reprinted by permission.

1 Sir Aure! Stein, Old Routes of Western Iran (London, 1940) pp. 376ff.
2 Stein, Old Routes, p. 377. An old landlord from Cheshm Gl told us that he remembered
a fat elderly American (i.e. someone who spoke English) who came to the area with an
Indian and his wife about 30 years ago after excavating at Dinkha. He also claimed that the
Indian found two vessels in one of the tumuli. It would seem that these vessels must actually
have come from one of the tepes sounded in the area by Stein (e.g. Stein, Old Routes, p. 377).
Tumuli G and H show unmistakable signs of having been excavated to a limited extent, and
they are probably the ones tested by Stein. However, two or three other tumuli also show
signs of excavation, infra.
3 Tumulus K, the eleventh recorded in 1968, was recognized as a tumulus by Christopher
Hamlin while he was planning the site.

10 chapter one
Figure 1. Tumuli I, E, and F, before excavation, right to left. Qalatgah lies to

the left of and above the cluster of trees in the right background.
Qalatgah, just slightly to the northeast (Figure 1).4 That site is known today
in the area for a pair of magnificent springs that gush from the rocks and
irrigate the fields below. All the tumuli are clearly visible from Qalatgah and
also from the modern Nagadeh-Ushnu road, which passes just below at the
base of the hills. The village of Cheshm Gl lies about half a kilometer to
the northwest of the tumuli.
In 1968 the Hasanlu Project, under the joint sponsorship of the Uni-
versity Museum of the University of Pennsylvania and The Metropolitan
4 Qalatgah means place of the fortress. The site was first visited by the author, Agha
Z. Rahmatian, and two members of the staff, Christopher and Carol Hamlin, on August 31,
1968. We were guided by a local landlord who promised to show us the place where an
inscribed stele was allegedly found in 1967. A major stretch of Urartian-type walls, Toprak-
kale-type pottery (highly polished red ware), and Iron III Iranian sherds were discovered. A
second trip a week later led to the discovery of an Urartian inscription on a building stone
and an Urartian stone stampcylinder seal. Collectively, the evidence suggests that Qalatgah
is an Urartian site.
the tumuli at s girdan: a preliminary report 11
Museum of Art, began a second campaign at Dinkha Tepe.5 Permission was

generously granted by the Iranian Department of Antiquities to conduct
a sondage at S Girdan. The aim of the sondage was to study the tumuli:
the way they were constructed, the type of tombs they contained, and any
evidence as to their date or the identity of the people who built them. No Ira-
nian tumuli of the type found at S Girdan have hitherto been excavated or
published. Any information provided by the sondage was certain, therefore,
to be of some significance and interest in the study of the ancient history of
Iran.
The area of the necropolis was first surveyed, and four tumuli were mea-
sured; the short time at our disposal precluded the measurement of the
other seven tumuli. The letters A to K were assigned to the tumuli for the
purpose of field identification; as they were selected for excavation, Roman
numerals were assigned. Seven of the tumuli lie roughly in a straight line
oriented northwest-southeast, extending over a distance of 600 meters (Fig-
ure 2). The four other tumuli lie to the north and northeast in no apparent
order. Tumulus I (i) is 750 meters northeast of Tumulus H; Tumulus K is iso-
lated about one kilometer east of Tumulus I (i), just off the modern road. A
large, a medium, and a small turnulus were chosen for examination. In two
of these tumuli, II and III (D and A), tombs were discovered and recorded,
while in the other, I (C), excavation had to be suspended before a tomb could
be located.
The main problem faced by the excavators was that the area surrounding
each tumulus was part of a cultivated pea field. The excavated earth from
all three tumuli had, therefore, to be deposited on their peripheries and
on the unexcavated sections (Figures 3, 4). Moreover, because the pea field
encroached on the tumuli, we were in no instance able to excavate the
original outer border.
5 For a report on the first campaign, see R.H. Dyson, jr., Dinkha Tepe, in Survey of
Excavations in Iran During 19651966, Iran 5 (1967) pp. 136ff.; Oscar White Muscarella,
Excavations at Dinkha Tepe, 1966, The Metropolitan Museum of Art Bulletin 25 (1966) pp. 16ff.
The 1968 campaign was codirected by the author and Professor Robert H. Dyson, Jr. The
author directed the work at S Girdan. He was ably assisted by Mr. Christopher L. Hamlin,
who planned the site and contoured Tumuli I, II, III, and E; Mrs. Carol Hamlin, who excavated
and planned Tumulus II; Miss Elizabeth Stone, who contoured Tumulus III; and Mr. Arthur
Smith, who was the photographer. Agha Z. Rahmatian was our most cooperative inspector,
who helped us in many ways. Agha Nozar Supheri was, as always, our right hand in all matters.
Our foreman, Beshir, deserves much credit and thanks for his intelligence, honesty, and help,
both to his Kurdish- and Turkish-speaking workers and to the excavation staff; through his
endeavors we were able to accomplish much work in little time.
12 chapter one
Figure 2. Plan of S Girdan showing Tumuli A to

H. Four tumuli, A, C, D, and E, have been planned.
Figure 3. Tumulus II, partially excavated, in the foreground;

Tumulus I in the background. Note the pea field and the dumps.
Figure 4. Tumulus II, during excavation, and Tumulus

E. Cheshm Gl may be seen in the upper right
background. Photograph taken from the top of Tumulus I.
14 chapter one
Figure 5. Plan of Tumulus I.
Figure 6. Section of test trench I, Tumulus I.

Figure 7. East-west section of the northwest quadrant, Tumulus I.
the tumuli at s girdan: a preliminary report
15
16 chapter one
Tumulus I
This is the largest tumulus at S Girdan (Figures 1, 2, 5): its height is 8.25
meters; its diameter, limited by an irrigation ditch and the pea field, mea-
sures about 6065 meters. Figures 6, 7, 11 show the present border of the
tumulus and the pea field. The tumulus was divided into quadrants using
compass directions as dividing lines. The northwest quadrant was further
subdivided into two sections, and the southern section was excavated. While
excavation progressed in the main cut, four narrow test trenches were exca-
vated around the tumulus (Figure 5).
The upper part of the tumulus fill was composed of a mixture of gravel
and earth, characterized by masses of small pebbles (1 in Figure 7). Below
this level the fill was composed of hard and firmly packed clay in combina-
tion with small pebbles, but not in the same quantity as in the upper level
(2 in Figure 7). After we reached this level, which continued until we sus-
pended digging, work proceeded very slowly.6 A color change was noticed
in the clay at a certain point, and there was also a lens of clay, but otherwise
nothing distinctive about the density and composition of the clay could be
detected (2A, 2B, and 2C in Figure 7). The color change may represent only
different sources for the clay.
After some days of digging, vertical cleavages were noticed in the north-
south scarp; the faces of some of these cleavages had a southeast, others
a northeast, orientation penetrating into the undug scarp. At first it was
assumed that these cleavages represented the drying and cracking of the
clay in the sun, but soon other cleavages were noticed in the surface of
the clay in the main cut. Work ceased for several days while these cleav-
age lines were cleaned and recorded. It was soon evident in the main cut
that a series of roughly concentric circular units were present (Figures 7,
8, 9, 10), and moreover, the cleavages joined neatly with some of the ver-
tical ones in the north-south scarp. Other cleavage lines were noticed in
addition to the circular ones: longitudinal lines that divided the circular
units into partitioned sections. Some of these partition-cleavage lines also
joined neatly with some in the north-south scarp, thus explaining the dif-
ferent angles observed there for the vertical cleavages. Examination of the
cleavages showed that there was a grain pattern on all the faces cleaned,
indicating that wood fences, stockade-like, had at one time served to con-
tain the clay and left their impressions upon it.
6 Each of the four strong pickmen found it necessary to strike the clay four to six times
before a section could be removed. The clay seems to have been packed in while still wet.
Figure 8. Circular cleavages in the surface of the main cut,

Tumulus I. Photograph taken from the top center of the tumulus.
Figure 9. Cleavages in the east-west section

and on the surface of the main cut, Tumulus I.
18 chapter one
Figure 10. Plan of cleavages in the main cut, Tumulus I.
Valuable information is thereby supplied to us relating to the techniques

employed in the erection of the tumulus. The builders first established
roughly circular areas by means of wood fencing, and they subsequently
subdivided these areas into irregular sections by means of wood partitions.
After the spaces so divided were filled with clay, it would seem that all wood
fences and partitions were removed, as no traces of such wood (other than
the imprints) were found. Except for the change of color recognized in the
scarp, no horizontal lines were visible; moreover, the vertical cleavages in
the scarp uniformly pass through the area of the color change. We must,
therefore, assume either that the fences and partitions were fairly high or
that they were continuously raised after a given amount and level of clay
had been deposited. We do not know at this stage of the excavation whether
or not the cleavage lines continue to the base of the tumulus.
A long trench was dug extending from the outer limit of the main cut
to a point slightly beyond the present periphery of the base of the tumulus
(Figures 5, 7). In this trench a neatly laid and compact sloping layer of small
stones, 1 to 3cm. in diameter, was encountered at irregular depths of from
40 to 60cm. below the surface. The stone layer was only one or two courses
thick and extended from a point near the present base up the slope for
a length of 10.5 meters, ending in an irregular line. At a depth of about
40cm. below the lower edge of the stones another layer of neatly laid small
stones was encountered, this time, however, laid horizontally (Figure 7,

lower right); where it begins and ends could not be established.
Each of the four test trenches excavated around the tumulus yielded the
same layer of small stones sloping up the sides (Figures 5, 6, 11). The depth
of the stones was irregular within each trench and also with respect to the
other trenches, and the length of the sloping stone layer in each trench was
not uniform. Gaps in the stones of test trench 4 could be explained either
as the result of stone robbing (though there is no evidence of this on the
surface), or by assuming that the builders began to run out of stones when
they reached this part of the tumulus and proceeded to pile the stones at
random. In one of the four test trenches, number 3, a horizontal course of
stones was found below the sloping course.
It is certain, therefore, that Tumulus I was encircled by a sloping revet-
ment of small stones neatly laid below the surface of the tumulus but at an
irregular height. This revetment was apparently built with the view to fur-
ther protecting the tomb, already covered with a mass of hard clay. The form
of the tumulus ultimately desired apparently did not develop until the time
when this revetment was covered with gravel, the last stage in the construc-
tion.
Although we did not discover a tomb, it may be safe to conclude, on
the evidence collected from Tumuli II and III (infra), that it exists in the
southwest quadrant and that it will be away from the center of the tumu-
lus.
Tumulus II
This is the smallest of the tumuli excavated in 1968, being about 47 cm.
in height and about 14 meters in preserved diameter (Figures 3, 4). The
mound was divided into quadrants, and the southwest one chosen for initial
excavation. The surface of the tumulus showed remains of recent hearths,
but no other features that might suggest disturbance were noticed.
The fill was very shallow and was composed of gravel and gray-brown
earth. At a depth of from 10cm, to 30cm., at different parts of the quadrant,
portions of a circular pile of rubble stones 10cm. to 30 cm. in diameter were
uncovered. In two areas there were gaps in the rock pile, and below the
larger gap we encountered the top of a well-built stone wall and a section
of another; this turned out to be the tomb. At this stage the rock pile was
completely excavated (Figure 12). It consists of a round mound of stones
several courses thick at the center and diminishing to one or two courses
20 chapter one
Figure 11. Stone revetment as seen in test trench 1, Tumulus I,

looking down the slope with the pea field in the background.
at the edges. The center of the rock pile was not under the center of the
tumulus but was actually some meters to the west. The tomb was placed at
the center of the rock pile.
Gaps in the rock pile noticed early in the excavation made it evident
that the tomb had been plundered: one gap in the southwest area seems
to indicate an abortive attempt; the large gap in the area directly over the
tomb represents a successful one. The stones scattered in this area, and
extending beyond the rock pile, at a higher level, represent debris from the
robbers trench, which was apparently dug from the east. This trench was
subsequently refilled with the same gravel and earth that covers the rest
of the tumulus. There was no evidence of the robbers trench in the north-
south balk despite the fact that there is a break in the line of stones. This
could indicate that the robbing occurred soon after the completion of the
tumulus and that the refill had consolidated with the undisturbed fill in
the course of time. Nothing on the surface of the tumulus gave any hint
that plundering had occurred, which further suggests that the robbery took
place in antiquity. Several of the other tumuli in the necropolis (I, III, F, G,
and H) have noticeable hollow depressions that indicate relatively recent
attempts at excavation. The depressions in two of these, F and G, represent
the soundings of Stein, according to some local inhabitants.
In the excavation of the tomb, the rubble stones of the overlying rock
pile were removed, and the south and east tomb walls were freed from
earth and rubble. The tomb (Figures 1316) is built of neatly cut large flat
stones, set into a thick mud mortar layer. The same mud used in the mortar
was also applied as a plaster to the outside walls of the tomb. The walls
form a rectangular structure one stone thick and about 1.5 meters by 3.1
meters. Around the top (except on the robbed south side) the flat stones
are bordered by additional large flat stones, which are themselves bordered
by rubble stones. On the three sides not disturbed by the robbers, the depth
of the tomb is preserved fully to a height of 1.2 meters. There is no evidence
to suggest of what material the roof was constructed.7
Below the walls to a depth of about one meter the tomb was filled with
rubble stones mixed with gravel and earth. Either these stones were thrown
in by the robbers or they fell in from the disturbed pile above. Under this
rubble fill, a layer of well-packed pebbles and some fist-sized stones were
encountered, apparently representing a deliberate packing. Among the
7 A wood roof could have been employed over the stone tomb, cf. M. Gimbutas, Bronze
Age Cultures in Central and Eastern Europe (The Hague, 1965) p. 284.
22 chapter one
Figure 12. Plan of Tumulus II showing

the tomb area and the rock pile overlay.
Figure 13. Tomb of Tumulus II during excavation.
Figure 14. Tomb of Tumulus II after excavation.

24 chapter one
Figure 15. Plan of the tomb, Tumulus II (Datum: +107).
Figure 16. Sections of the tomb, north and west walls, Tumulus II.
pebbles were two thin lenses of fine gray ash whose origin remains a mystery.
Under the packing was the tomb floor proper, constructed of large flat stones
of the same type used in the walls. The center of the tomb floor contained
no slabs but only a smooth and hard gravel surface spotted with red stains,
samples of which were collected. A test trench dug 25 cm. into this surface
demonstrated that it was virgin soil. It may well be that the pebble packing
rather than the partially slabbed floor served as the surface on which the
dead person was placed, and that the slabs were missing in the original tomb
construction.
Attempts were made to find out whether a tomb pit had been dug, but no
evidence of one was found. Sections were cut at both the north and south
ends of the tomb area, and no pit lines could be recognized. The present
surface of the valley is 2 meters above the stone floor of the tomb, and it
would indeed be possible to assume that a pit had been dug for the tomb.
Perhaps, since the gravel used as fill was of very much the same consistency
as the neighboring virgin soil, the outlines of the pit have been obscured.
The contents of the tomb consisted of a few small bone fragments in very
poor condition, found in the northwest corner at a depth of 40 cm. from the
top of the wall, and a small, nondescript disc-shaped shell bead. A small,
coarse sherd was found below the rock pile overlying the northeast corner
of the tomb, but it unfortunately yields no information.
Three sherds, each incised with part of a triangle, were found in three
areas of the tumulus fill, in each case just beyond the circular rock pile. We
will return to a discussion of these sherds shortly.
It may be seen from the plans and photographs that Tumulus II has not
been completely excavated. It would therefore be premature to arrive at a
negative conclusion concerning the presence of a circular stone revetment,
as was excavated in both Tumulus I and Tumulus III. The tomb area and
the overlying circular rock pile have been cleared, but not the outer areas of
the tumulus. Perhaps during a future season conditions in the pea field will
allow a test trench to be dug in a search for a revetment.
Tumulus III
Tumulus III was selected for excavation because it represents a medium-

sized example, being 3 meters in height and about 35 meters in preserved
diameter (Figure 17). Whereas Tumuli I and II were apparently never cul-
tivated, Tumulus III had been plowed and therefore blended into the sur-
rounding pea field. After the tumulus was divided into quadrants, the
26 chapter one
Figure 17. Plan of Tumulus III.
Figure 18. East-west section of the southeast quadrant, Turnulus III.

southeast section was decided upon for initial excavation. The fill through-
out was solid clay mixed with a few pebbles. Aside from the upper area,
which had been softened by plowing, this clay was quite hard and compactly
laid down. No horizontal lines were visible in the balk, nor were there any
cleavages such as those recognized in Tumulus I (Figure 18).
Approximately 50cm. below the surface around the outer perimeter of
the tumulus, a section of a sloping ring of rubble stones one or two courses
thick and of varying sizes was uncovered (Figures 18, 19, 20). The ring is about
5.75 to 6 meters in width (measured horizontally) and seems to have served
as a revetment. Although many stones are missing as a result of plowing,
there is no evidence of plundering. A test trench was cut into the north slope
of the tumulus (Figure 17) and a sloping stone surface thereby uncovered,
indicating that the revetment encircled the tumulus in the same manner as
recorded in Tumulus I.
Some 2.5 meters below the center of the tumulus a small section of a
rubble rock pile was encountered. The outer border of the pile was one stone
thick, and the pile increased in depth toward its center, thus forming a low
mound; the outer edge, which was curved, indicated that the pile was round
in form (Figures 19, 21). The rock pile rested on a well-made floor of clay,
smooth and hard, only part of which could be cleared. Since it was evident
that the center of the rock pile was located in the southwest quadrant, a
trench extension was made in that direction; another extension was made
to the north to expose more of the rock pile and to allow extra room for
excavating the tomb.
A hollow dome in the clay over a depression in the stones 1.1 meters deep
indicated clearly that the rock pile had collapsed into the tomb chamber.
When this area was cleared, the top of a grave pit completely filled with
fallen rubble stones was revealed. Powdery remains and small fragments
of wood, which apparently belonged to the now decayed roof, were also
recovered mixed with the rubble. No rubble stones found in the area could
have served as a roof for the grave pit.
The grave pit was a neatly excavated rectangular area 2 meters wide, 3.5
meters long, and 1.2 meters deep. The walls sloped outward slightly and were
coated with a thin mud plaster. The level gravel floor of the pit was carefully
covered with a layer of small pebbles to a depth of 78 cm. (Figure 22).
Excavation into the gravel floor showed it to be virgin soil.
On the pebble surface was found the badly crushed skeleton of an adult
male (Figures 23, 24). It lay on its side with the head facing southeast and
the legs drawn up in a contracted position. The left arm was awkwardly
positioned under the right arm, and the back was twisted. The original
28 chapter one
Figure 19. Plan of Tumulus III showing excavated areas:

circular stone revetment, inner rock pile, and tomb area.
Figure 20. The excavated southeast quadrant of Tumulus III

with the outer stone revetment in the foreground
and the inner rock pile just above it in the center.
Figure 21. Inner rock pile, Tumulus III.

30 chapter one
Figure 22. South and west sections of the tomb pit, Tumulus III.
Figure 23. Plan of the tomb, Tumulus III.

Figure 24. Skeleton on the floor of the tomb pit, Tumulus III.
position of the skeleton may have been distorted by the rubble collapse, but
the placement of the arms seems surely to have been original. All bones of
the skeleton were completely covered with a deep red color, specimens of
which have been taken for analysis.
The following objects were found on the pebble floor beside the skeleton
and clustered in an area to the west and northwest of the head (Figures 23,
24):
1. A perfectly preserved whetstone-like object terminating in a felines
head. The stone is finely grained and very smooth so that if it is a
whetstone it does not appear to have been used; perhaps it served
as a baton or scepter. Length 37cm., diameter 3.2 cm. (Figures 25, 26,
27).
2. A very fragmentary silver drinking vessel, the metal of which was in
excellent condition when found (Figure 28 for a reconstructed draw-
ing; Figure 23, 2 in plan).
32 chapter one
Figures 25, 26, 27. Stone whetstone-scepter,

Tumulus III. Iran Bastan Museum, Tehran.
3. A bronze knife blade with the remains of a plaited material adhering

to one side. Length 12.8cm., greatest width 1.8 cm.
4. A bronze celt, also with the remains of a plaited material adhering to
one side. Both the knife and the celt may have been resting on this
material, as fragments of it were recovered underneath both objects.
Length 13.4cm., width 5.13.5cm. (Figure 29).
5. Two long silver rods, very fragmentary. They were made by rolling silver
plate in the manner of a scroll.
6. Many small beads of gold, stone, and paste. They are all plain and are
round or rectangular in shape. The beads were recovered near the right
hand; none were found near the neck (Figure 23, 6 in plan).
Picked up in the tumulus fill were a few sherds of coarse ware, a few sherds
decorated with incised wavy lines of second-millennium type, and three
small and fragmentary human bones. All these objects could have been
inadvertently deposited along with the clay.
Figure 28. Reconstructed drawing of a silver

vessel, Tumulus III. Iran Bastan Museum, Tehran.
Figure 29. Bronze celt blade, Tumulus III. Iran Bastan Museum, Tehran.
34 chapter one
The building of the tumulus may be reconstructed as follows: An area in

the field was leveled and smoothed to a neat, hard surface. Into this surface
a rectangular pit was dug, floored with pebbles, and lined with plaster. After
the deposition of the body and of the objects, wood beams or logs were
placed over the pit, and a carefully laid mound of stones was placed over
the closed tomb, which was kept under the center of the rock pile. Following
this stage came the laying down of the clay, in which the center of the rock
pile and tomb were kept away from the center of the tumulus. No stones
or wood remains that might have served as a marker for the center of the
tumulus were found.8 At a certain stage, near the completion of the tumulus,
a sloping, encircling stone revetment was built. Following this stage, more
clay was dumped in order to cover the revetment and to create the shape
desired for the tumulus. None of the evidence suggests that wood fences or
partitions were employed in the construction.
Summary
Tumuli II and III share certain features: each of their tombs was situated
away from the center of the overlying tumulus; each tomb was built into a
pit (not absolutely certain for II); each tomb had a pebble floor; and each
tomb was covered by a mound of rubble stones. Until a trench is cut into
the outer area of Tumulus II, we are not in a position to conclude that it
had a stone revetment like Tumulus III, but this is very probable. The major
difference between the two tumuli, aside from size, is that II contained a
well-built stone tomb whereas III contained only a simple pit as the grave
chamber. Whether this fact may be interpreted as reflecting a difference in
wealth between those who were buried in the tomb or a chronological gap
between the erection of the tumuli cannot yet be established.9
A feature Tumulus I shares with Tumulus III is the stone revetment.
With respect to the unique technique used in the construction of Tumulus I
(wood fencing), one may presume that the relatively small size of Tumuli II
and III precluded such an elaborate system.
8 Such markers were found within some Phrygian tumuli: G. and A. Koerte, Gordion
Ergebnisse der Ausgrabung im Jahre 1900 ( Jahrbuch des deutschen archologischen Institut,
Ergnzungsheft V; Berlin, 1904) p. 39; R.S. Young, Gordion 1956: Preliminary Report, Ameri-
can Journal of Archaeology 61 (1957) p. 325 (Tumulus P); idem, The Gordion Campaign of 1959:
Preliminary Report, AJA 64 (1960) p. 228 (Tumulus W).
9 The low height of Tumulus II might have resulted to some extent from the plundering
activity, but there is no conclusive evidence that it was ever as high as Tumulus III.
Tumuli whose construction exactly parallels that of the tumuli at S

Girdan do not come readily to mind. Particular features of the construetion,
however, are paralleled in various areas of the ancient world, extending from
England to Russia, and possibly beyond.10
The placement of the tomb away from the center of the tumulus is a
characteristic feature of Phrygian and Lydian tumuli in Anatolia beginning
in the eighth century bc and continuing for several centuries.
Phrygian tumuli excavated at Ankara, Gordion, and Kerkenes Dagh, with
rare exceptions, have a grave chamber off-center.11 The Phrygian tumuli
usually also have a grave pit into which a tomb was placed; they sometimes
have a pebble floor; and they almost always have either rubble stones placed
around the tomb in the pit or, more commonly, a rock pile covering the tomb
pit. In at least one Phrygian tumulus, one of those excavated by Makridi at
Ankara, there is evidence that fences or partitions were erected to help in
the orderly dumping of earth fill.12 In the case of this particular example, the
partition walls were constructed of stones and were left in place as the fill
rose (cf. Salamis, infra).
None of the Phrygian tumuli have an outer stone revetment, and this
seems to be the only important structural difference between these tumuli
and those at S Girdan. Another difference is that in Phrygian tumuli the
10 A fragment from the Stele of the Vultures of Eannatum, now in the Louvre, has a scene
that could represent the erection of a tumulus over a mass burial. Two men carry earth up a
ladder or slope in order to cover a group of dead men, the defeated enemy. The fact that they
are climbing seems to preclude the suggestion that we are witnessing a regular inhumation
burial. However, there is at present no known tumulus burial from the Mesopotamian area of
this time, or indeed later. For a drawing of the fragment and a suggestion that the scene does
represent the erection of a tumulus, see G. Perrot and C. Chipiez, A History of Art in Chaldaea
and Assyria (New York, 1884) pp. 177 ff., fig. 93.
11 Ankara: M. Schede in Archologischer Anzeiger 1930, col. 480; T. zg: and M. Akok,
Die Ausgrabungen an zwei Tumuli auf dem Mausoleumshgel bei Ankara, Belleten 11 (1947)
p. 59, where the excavators statement that the tomb of Tumulus 1 was directly under the
Gipfel seems contradicted by fig. 5; on p. 69 they state that the tomb of Tumulus 2 was
unter der Mitte des Hgels Gordion: Koerte, Gordion, pp. 99, 105, 139ff. (Tumuli IV, II, V);
the Koertes specifically state, p. 40, that the tomb was under the center of Tumulus III, which
appears to be an exception at Gordion; R.S. Young in various preliminary reports: AJA 61 (1957)
p. 325 (Tumulus P); AJA 64 (1960) p. 228 (Tumulus W); AJA 62 (1958) p. 147 (Tumulus MM);
Bulletin of the University Museum 16 (1951) p. 11, pl. V (Tumulus G); Archaeology 3 (1950) p. 200,
fig. 7 (Tumulus B); AJA 70 (1966) pp. 267 ff. (Tumuli X and Y). Kerkenes Dagh: E. Schmidt, Test
Excavations in the City of Kerkenes Dagh, American Journal of Semitic Languages 45 (1929)
pp. 250 ff.
12 H.H. von der Osten, Explorations in Central Anatolia Season of 1926, Oriental Institute
Publication, V (Chicago, 1929) pp. 48 ff., fig. 78; M. Schede in AA 1930, cols. 479ff., fig. 23;
M. Forrer in Mitteilungen der Deutschen Orient Gesellschaft 65 (1927) abb. 20 with caption.
36 chapter one
tombs are usually constructed of wood, but this may only reflect ecological
differences between Anatolia and western Iran.
Several Lydian tumuli excavated in the region of Sardis also have the
tomb placed off-center, a feature, incidentally, that one may interpret as an
example of Lydias cultural dependence on Phrygia.13 In some tumuli there is
also a pile of rubble stones placed over the tomb. While there is no evidence
of a stone rubble revetment in the tumuli at Sardis, the elaborate stone wall
found in the large tumulus called Karniyarik Tepe may actually be related in
some manner to the type known at S Girdan.
Still another area where there are tombs placed off-center is on the
island of Cyprus at the necropolis at Salamis, recently excavated by V. Kara-
georghis.14 One of the tumuli, called Tomb 77, is a fourth-century bc ceno-
taph. The pyre, with its contents and covering rock pile, was excavated intact
only because it was missed by grave robbers who had tunneled straight for
the center of the tumulus and thereby missed their goal. In addition to these
parallels with S Girdan, the tomb off-center and the overlying rock pile,
there is another feature of some importance: thin stone rubble walls were
found that radiated out from the center of the tumulus, dividing the area
into sections in order to facilitate the orderly dumping in of the earth fill.
This employment of stone partitions was also recorded in Tumulus 3 of the
same necropolis.
Some European tumuli also present interesting parallels to those at S
Girdan. Tumuli of the second millennium bc excavated in the western
Ukraine, the Baltic area, and central Europe often have a tomb pit that is
covered with a pile of rubble stones. In addition, some have a simple stone
ring encircling the tumulus at the base.15 These stone rings do not seem to
have functioned as a revetment, as we suggested stone rings did at S Girdan,
13 Perrot and Chipiez, History of Art in Phrygia, Lydia, Caria, and Lycia (New York, 1892)
p. 262, fig. 159, p. 270; H.C. Butler, Sardis, I (Leyden, 1922) p. 10; G.M.A. Hanfmann, The Fifth
Campaign at Sardis (1962), Bulletin of the American Schools of Oriental Research 170 (1963)
p. 56, fig. 39; idem, The Ninth Campaign at Sardis (1966), BASOR 186 (1967) p. 39, fig. 25,
pp. 42 ff.; on p. 47 there is mention of a tumulus called Cambaz Tepe that has a tomb placed
under the center. For a brief discussion of Phrygian influence on Lydia see Oscar White
Muscarella, Phrygian Fibulae from Gordion (London, 1967) p. 44, notes 30, 31.
14 V. Karagcorghis, Chronique des Fouilles en Chypre, Bulletin de Correspondence Hel-
lnique 90 (1966) p. 377 (Tomb 77); BCH 89 (1965) p. 283 (Tomb 3, seventh century bc).
15 Gimbutas, Bronze Age Cultures, pp. 285, 308, 319ff., 420, 460, figs. 190, 212, 219, 273, 301.
Parenthetically I would mention in this context the tumulus covering the House of Tiles
at Lerna. A circle of stones surrounded thc earth mound, and a layer of stones covered its
surface. No burial was found, but the construction is certainly similar to that employed in
grave tumuli: Hesperia 25 (1956) p. 150, fig. 3, pp. 165 ff., fig. 5.
but the same general ideaan encircling of the tumulus with stones
seems to be in evidence. And it is this feature that particularly relates the
S Girdan tumuli to those known in Europe. An important difference may
be seen in the fact that it was normal for a European tumulus to have a tomb
built directly under its center.
One tumulus known to me from England that has a feature recognized
at S Girdan is a long barrow at Skendleby in Lincolnshire. Its excavation
produced evidence that upright wood posts and fences were employed in
its construction. The system of partitioning, or dividing areas into units, was
also recorded.16 This technique is the same as that employed in Tumulus I,
and, as already discussed, in tumuli from Ankara and Cyprus.
Tumuli in the Altai area do not present direct parallels with those at
S Girdan in that many are actually rock cairns rather than earth tumuli.
The rock pile covering the tomb pit was itself the tumulus and was usually
centered over the tomb below.17
An outer ring of stones surrounding a centrally positioned tomb is fairly
common in the Caucasus region.18 These stones, however, form a simple ring
and not a revetment, and as such may be more closely related to European
tumuli. I am told that some of the tumuli ruthlessly plundered in the Ardebil
area in recent years have an outer ring of stones like those in the Caucasus.
The objects recovered from the rubble-filled tomb at S Girdan do not
yield as much information about culture and chronology as one would wish;
especially lacking is pottery. Nevertheless, certain general comments may
be set down. The silver drinking vessel, although badly crushed, can be
partially reconstructed on paper (Figure 28) and has several parallels at
Ziwiye belonging to the Iron III period, there not earlier than the late eighth
century bc and continuing through the late seventh century bc19 The knife
and celt, however, are not distinctive enough to allow them to be placed
chronologically with any certainty.
16 C.W. Phillips, The Excavation of the Giants Hills Long Barrow, Skendleby, Lincoln-
shire, Archaeologia 85 (1936) pp. 60 ff., fig. 7 on p. 61, and pl. XXI, fig. 2.
17 M.R. Griaznov and E.A. Golomshtok, The Pazirik Burial at Altai, AJA 37 (1933) pp. 32ff.,
fig. 1; K. Jettmar, Art of the Steppes (New York, 1964) pp. 120ff., fig. 105.
18 J. de Morgan, Mission Scientifique en Perse (Paris, 1896) p. 43, fig. 45; H. de Morgan,
Recherches au Talyche Persan, Dlgation en Perse, Mmoires 8 (Paris, 1905) pp. 256ff.,
figs. 339342, p. 260, fig. 346, p. 262, figs. 347, 348, etc. For similar types of tombs in Italy see
A. Minto, Marsiliana DAlbagna (Florence, 1921) pp. 22 ff., 30ff., fig. 2, pl. VI.
19 T. Cuyler Young, Jr., A Comparative Ceramic Chronology for Western Iran, 1500500bc,
Iran 3 (1965) P. 60, fig. 4, no. 10. There are examples from Ziwiye still unpublished.
38 chapter one
The most important object from the tomb, and the only one completely
preserved, is the feline-headed whetstone-scepter. The head of the feline is
stylized and simple in execution, and seems to be pre-Achaemenid in style.
Whetstones, often with detachable metal animal heads, are known from the
Achaemenian period and earlier.20 A few whetstones of a similar but smaller
type are reported from the Minusinsk area in Russia. These have animal
heads and were made in one piece. Tallgren states that they are difficult to
date.21 I would tentatively suggest a pre-Achaemenid date for the whetstone-
scepter found in Tumulus III.
The few sherds with incised triangles found in the fill of Tumulus II
(Figure 30) may fit within the Iron III period, perhaps late eighth to early
sixth century bc Painted and incised triangles were common motifs in that
period (hence Iron IIIs original field name Triangle Ware Period) at such
sites as Hasanlu, Susa, Ziwiye, and Zendan, and at sites in the Caucasus.22
Recently there have appeared on the antiquities market vessels with incised
triangular decoration allegedly coming from northwest Iran (Figure 31);
these also seem to belong to the Iron III period.23
Sherds and other material found in the fill of a tumulus do not date its
construction except in the form of an ante quem non date,24 that is to say, the
objects may be interpreted as either contemporary with the erection of the
tumulus (a workman scattered a pot he accidentally broke) or earlier than
20 7000 Ans d Art en Iran (Paris, 1961) no. 689; R. Ghirshman, The Arts of Ancient Iran
(New York, 1964) p. 67, figs. 8436; A. Godard, Bronzes du Luristan (Paris, 1931) pls. XI, XII, an
example from Susa and others from Luristan. In this context compare a door socket made
of a finely grained stone with a stylized rams head at one end, found at Hasanlu, dated to the
ninth century, R.H. Dyson, Treasures From Hasanlu , Illustrated London News, Sept. 30,
1961, p. 536, fig. 12. The scepters discussed by D. Berciu, A Zoomorphic Sceptre Discovered
in the Peoples Republic of Bulgaria , Dacia 6 (1962) pp. 397ff., may be related to but are not
of the same type as the one from S Girdan. For this last reference I wish to thank Professor
T. Sulimirski.
21 A.M. Tallgren, Some North Eurasian Sculptures, Eurasia Septentrionalis Antiqua 12
(1938) p. 119, fig. 8a, b, p. 126.

22 R. Ghirshman, Village Perse-Achmnid, MDP 36 (Paris, 1954) pl. XXXI, GS. 863, pl.
XXXIII, GS. 3; Young, Comparative Ceramic Chronology, pp. 68ff., p. 56, fig. 2; R.H. Dyson,
Jr., Problems of Protohistoric Iran as Seen from Hasanlu, Journal of Near Eastern Studies 24
(1965) pp. 204 ff., figs. 7, 9, 10, 11, 13; idem, Iran, 1956, Bull. Univ. Mus. 21 (1957) p. 36, fig. 27;
W. Kleiss and R.M. Boehmer, Die Ausgrabungen aufdem Zendan-i-Suleiman, AA 1965, cols.
763 ff., figs. 7476; J. de Morgan, Mission Scientifique au Caucase (Paris, 1889) p. 148, fig. 155,
p. 151, fig. 162, p. 155, fig. 170. (I am not unaware of the fact that the sherds from Tumulus II
could be second millennium in date.)
23 See also Trsors de l Ancien Iran, Muse Rath (Geneva, Ig66) no. 672, fig. 64. Many others
are to be seen in dealers shops.

24 Muscarella, Phrygian Fibulae, p. 7; Karageorghis, BGH go (1966) p. 377.
Figure 30. Sherds from the fill of Tumulus II.

40 chapter one
Figure 31. Vessel with incised triangular decoration, Iran, VIIVI

century bc. The Metropolitan Museum of Art, Rogers Fund, 66.202.
the erection of the tumulus (the material was inadvertantly scooped up by

workers while they gathered clay for the fill). If I am correct in attributing
the sherds to the Iron III period, then Tumulus II and, I would suggest, also
the others are either Iron III or later in date. The S Girdan tumuli remind
us strongly of the Phrygian and Lydian examples of the eighth through the
sixth centuries bc, the silver vessel has Iron III parallels, and the whetstone
appears to be pre-Achaemenian in stylethese three factors do suggest that
we would be on the right track in tentatively dating the construction of the
tumuli as seventh or sixth century bc.
A preliminary report cannot be more definite, and C-14 samples are still
to be collected and evaluated. Another seasons work might supply the
much-desired pottery needed to arrive at a stronger conclusion. One of
the problems, in addition to chronology, that requires further research is
that of the location in the Ushnu valley of the settlement occupied by the
people who buried their dead at S Girdan. Another related problem, one of
some importance, is concerned with the cultural identification of the tumuli
builders: were they, for example, Medes or Scythians, or still another people?
chapter two
THE TUMULI AT S GIRDAN: SECOND REPORT*
In July of 1970 the Hasanlu Project, under the joint sponsorship of The
Metropolitan Museum of Art and the University of Pennsylvania, began its
second campaign at S Girdan, situated in the Ushnu valley in northwestern
Iran.1 It will be recalled that S Girdan is a cemetery consisting of eleven
tumuli of various sizes near the modern village of Cheshm Gl and below
the recently discovered Urartian site of Qalatgah.2 The campaign of 1968 had
been basically a survey resulting in the partial clearing of the largest tumulus
there, designated I, the excavation of a small plundered tumulus, II, and the
excavation of an intact tumulus, III.
The aim of the second campaign was to excavate three additional tumuli
and to complete the excavation of Tumulus I with the view to learning some-
thing about the culture and chronology of the people buried in the cemetery,
since information of this sort had not been firmly established in the first
season. Our season did not begin until two weeks later than planned, and
we were therefore not able to complete the clearing of Tumulus I. We were,
however, able to excavate three of the other tumuli in the area, called IV, V,
and VI.3 On the plan published in S Girdan I, fig. 2, these tumuli are labeled
E, F, and G.

1 Oscar White Muscarella, The Tumuli at S Girdan: A Preliminary Report, Metropolitan
Museum Journal 2 (1969) pp. 5 ff. (hereafter S Girdan I). Credits for the drawings in this first
report are as follows: figs. 12, 15, 16 are by Carol Hamlin; figs. 6, 7, 10, 17, 18, 19, 22, 23 are by the
author; the tracings from field notes and subsequent inking are by Maude de Schauensee. In
the present report initials are included with the drawings; Marie Miller did the inking of the
drawings. I wish to thank all for their help and cooperation in these undertakings.
2 S Girdan I, p. 5, note 4; a discussion of Qalatgah by the author will appear in a future
issue of Expedition. For a discussion of other Urartian sites in Iranian Azerbaijan see W. Kleiss,
Bericht ber zwei Erkundungsfahrten in Nordwest Iran, Archaeologische Mitteilungen aus
Iran 2 (1969) pp. 20 ff.; and W. Kleiss, Urartische Pltze in Iranisch-Azerbaijan, Istanbuler
Mitteilungen 18 (1968) pp. 1 ff.
3 The season began on July 17 and ended on August 27. The staff consisted of the author
as director and Michael Nimtz (University of Pennsylvania), Karen Rubinson (Columbia Uni-
versity), and Betty Schlossman (Briarcliff College) as archaeologists; Robert Lewis surveyed
and oriented the tumuli. Agha Nozar Sepheri was the able assistant to the director, and Agha
44 chapter two
Tumulus I
Work was concentrated on the completion of the wedge-shaped trench

begun in 1968 in the southern part of the northwest quadrant (S Girdan I,
figs. 3, 5). Although we assumed that the tomb would not be found here, it
was thought best to finish this area in order both to get a complete section
of the tumulus and to uncover at least part of the rubble stones that we
assumed would be overlying the tomb. If the rubble stones could be cleared
and measured, we would presumably get information about the position of
the tomb.
The trench was cleared down to the anticipated rubble stones (Figure 1),
and a completed section of the trench was made (Figure 2). At a depth of 7
meters from the top of the tumulus we encountered soft clay (as opposed
to the generally compact clay above) and then the rubble stones within
50cm. These stones are fairly large, averaging between 20 and 65 cm. in
diameter, and clearly form a roughly circular mass; they were laid down
in several layers and mound toward the center. We cleared 2.70 meters
of the rubble mass measuring out from the undug south balk. It seems
certain that the tomb lies some distance to the south within the undug
area.
The rubble mass was covered with about 50 cm. of soft clay, as stated, and
over this began the mass of hard compact clay continuing for about 5 meters;
earth and gravel were dumped above. The cleavage lines recognized in 1968
continued down to the top of the rubble mass (Figure 2; S Girdan I, pp. 711).
There were no visible signs that the cleared section of the rubble mass had
been tampered with, and it would appear that the tomb remains unplun-
dered. Its excavation will have to await a future season.
In S Girdan I, pp. 1113 and fig. 5, it was recorded that 40 cm. below
the sloping revetment stones in the main trench and 50 cm. below the
revetment stones in test trench 3 (in the eastern part of the tumulus)4
Ardeshir Ferzanegan was the representative of the Iranian Archaeological Service. My aim in
both S Girdan I and the present report is to publish the archaeological results as quickly as
possible. I therefore do not claim to have exhausted all the evidence available for purposes
of seeking comparisons and relationships.
4 Unfortunately, in S Girdan I, p. 9, fig. 5, the lower horizontal course was inadvertently
not recorded correctly in test trench 3. It is interesting to note that when we redug the test
trenches filled in by us in 1968 we found the gravel and earth to be hard and without any
indication that they had been dug two years previously.
the tumuli at s girdan: second report 45
Figure 1. Rubble stones of Tumulus I. Note cleavage line at right.

46 chapter two
Figure 2. East-west section of the northwest quadrant, Tumulus I.
a horizontal course of stones was found; this course did not appear in
the three other test trenches dug. In 1970 we checked this information by
redigging test trenches I and 4: no horizontal course was found. Why this
course only occurs in the long trench and in one test trench is not known.
Perhaps certain parts of the ground needed leveling.
Tumulus II
An attempt was made to dig a trench outward from the area cleared in 1968
in order to confirm that there was a stone revetment (S Girdan I, p. 16).
Unfortunately, the landlord refused us permission although we promised to
refill the trench. There can be little doubt, however, that this tumulus had a
revetment since all the other tumuli excavated had such a feature.
Tumulus IV
Tumulus IV is the second largest in the S Girdan cemetery; it is about 7.5

to 8 meters in height and about 52 to 58 meters in diameter (Figure 3).
This tumulus, like all the others, is delimited by an irrigation ditch and
cultivated fields. Its surface was covered by prickly weeds and was not under
cultivation (S Girdan I, fig. 2, E on plan, figs. 1, 4).
The tumulus was divided into quadrants, with true north (14 14' east of
magnetic north) as orientation (n.b.: all the tumuli excavated in 1968 were
divided into quadrants on a magnetic north orientation; those excavated in
Figure 3. Plan of Tumulus IV.
1970 were given a true north orientation).5 Excavation began in the upper
part of the southwest quadrant within an arc forming a wedge-shaped area
extending 11.15 meters from the top of the mound.
Beginning about 1.50 meters west of the center of the mound, and just
below the surface, a closely packed mass of small stones10 to 20 cm. in
diameterwas uncovered (Figure 4). The mass was approximately 3 by 6
meters in area and did not form any recognizable plan; it extended partly
into the northwest quadrant and was one to two layers thick. A coarse,
carinated bowl, dark gray in color, and showing evidence of burning, was
found nestled within the stones on the southeastern edge; a stone was found
inside the bowl (Figure 27). I will return to this bowl later.
5 This change was inadvertent. I assumed mistakenly that our surveyor in 1970 would
use magnetic north as orientation. When I discovered the change, it was too late to make a
correction.
48 chapter two
Figure 4. Plan of stones below the surface at the top of Tumulus IV.
The stones were removed, and excavation continued to a depth of 2.5

meters, whereupon the area of excavation was limited to two trenches at
right angles to each other, along the north-south and east-west sides of the
quadrant; the trenches were respectively 1.75 meters and 1.50 meters wide
(Figure 3).
A short time after excavation started, an area of earth different in
color from the surrounding earth was noted in the east-west section
beginning just below the aforementioned stones. It soon became clear that
the area was a narrow shaft, now filled in, that at one time penetrated
into the tumulus (Figures 5, 7). At a depth of about 3.50 meters a hori-
zontal tunnel was encountered extending from the shaft southward 1.75
Figure 5. East-west section of the southwest quadrant,

Tumulus IV. The tomb is restored for convenience.
Figure 6. North-south section of the southwest quadrant, Tumulus IV.

50 chapter two
Figure 7. Part of the east-west and north-south sections of the southwest

quadrant, Tumulus IV. Part of the tunnel may be seen to the left and
below the metal tray; the western part of the tunnel has been removed.
Figure 8. Tomb chamber in cavity of Tumulus IV.

The ring of stones is to be seen in the foreground.
meters, then turning westward until it disappeared into the undug balk
(the west balk of the north-south trench). The tunnel was more than one
meter in height, although we could not measure it exactly because its course
was directly over a hollow cavity and it was not considered safe to work
there.
The cavity began at a depth of about 5.50 meters, just below the tunnel.
When the loose fill at the bottom of the cavity was cleared, the upper part of
the tomb was exposed; the tomb chamber itself was completely filled with
earth. The cavity extended over the whole area of the tomb (Figure 8), and at
the southwest corner it became a tunnel that continued southwestward and
upward from the tomb, beginning at a place where the latter was damaged,
several stones having been torn away (Figures 9, 10). This part of the tunnel-
cavity could safely be explored only superficially, but loose slabs of stone
were seen there, slabs that certainly had been torn away from the walls of
the tomb chamber.
The relationship of the cavity over the tomb and the vertical shaft and
horizontal tunnel seems quite clear. Whoever dug the shaft, i.e., the tomb
robbers, began it in an attempt to reach the tomb. (Because the shaft is so
narrow in section, it may be assumed that we cut into it near its perimeter
rather than at its widest part.) At a depth of 3.50 meters they decided to dig
a tunnel, first going south, then west, and finally north, moving downward
until they reached the tomb at its southwest corner. It seems plausible
to assume that the robbers had a general idea where to find the tomb
but were uncertain about its exact position. The clay over the tomb had
been removed laboriously through the tunnel and up the shaft, work that
must have been slow and hard, and we may assume that there were many
helpers. The firmness of the clay kept the cavity and part of the tunnel
intact, but the roof of the tunnel where it left the shaft had collapsed; the
vertical shaft also filled up with earth and stones in the course of time
in fact, may have been deliberately filled in so as to cover any traces of the
robbery.
The tomb lies in the western part of the tumulus, mostly in the southwest
quadrant, partly in the northwest quadrant (using either true or magnetic
north as orientation). It was placed so that, but for part of the short east
wall, it was away from the center of the tumulus (infra) (Figure 10). The
cavity created by the robbers digging activities extended over the entire
tomb and cleared part of the upper surface of all the walls. It took us
several days to clean out the earth fill in the tomb, mainly because, as we
approached the bottom, we encountered thick, wet mud, the water table
being at hand.
52 chapter two
Figure 9. The western and part of the northern and southern walls of the
tomb, Tumulus IV. Note the robbers tunnel and entry at the left.
Figure 10. Plan of the tomb and outer stone ring, Tumulus IV.
The tomb is a well-made structure, rectangular in plan, with internal

measurements of 5.85 by 2.27 meters and an east-west orientation. It was
carefully constructed of rectangular slabs of stone laid in thirteen courses,
with thick mud mortar layers separating the slabs (Figures 11, 12). The slabs
vary in size, averaging 60 to 80cm. in length and 8 to 15 cm. in width; a
few slabs are shorter in length, while others are as long as 1.50 meters. The
mortar thickness varies from 4 to 20cm., and the slabs and mortar layers do
not always coincide from course to course. The corners of the tomb were
built at right angles and interlock, some stones of one wall thrust into the
other, locked in place by the upper and lower stones; this does not occur in
a regular fashion (Figure 12).6
Large stone pebbles 5 to 15cm. in diameter were found in the mud and
cleared out. We are not sure whether or not these stones represent the floor
of the tomb: it was not possible to determine any order with respect to the
stones because of the mud. Excavation stopped at a point just below the
lowest stone course of the tomb, but we encountered no clear indication
that there was a floor. The depth of the tomb, based on a measurement of
the walls, is about 1.60 to 1.70 meters.
6 Although not too clear from the drawing in S Girdan I, fig. 16, the corners of the walls
of the tomb of Tumulus II were made in the same way.

54 chapter two
Figure 11. View of the tomb facing east, Tumulus IV.
Figure 12. Section of the east wall of the tomb, Tumulus IV.
In the southwest corner of the tomb five courses of stone were missing for
about 1 meter to the east; in the western wall two courses were missing for
about 40cm. to the north (Figures 9, 10). This destruction occurred when the
robbers entered the tomb from their tunnel, where, as already stated, some
of the slabs were to be seen.
Although the tomb had been plundered, the robbers left some objects
because of either haste or carelessness. No skeleton was found, but we did
recover a few fragments of human bone, all showing definite red coloration;
some of the pebbles from the tomb also had this red color. A small fragment
of a smoky-clear obsidian blade was found in the fill above the tomb, and
one of the stone pebbles removed from the mud was a red chert core
from which blades had been chipped. Presumably this was not part of
the tomb equipment but just another rubble stone. If we are correct in
this observation, the stone must have come from a local field and suggests
that a neolithic or earlier settlement existed in the area. A few coarse,
nondiagnostic sherds and a few scraps of nondescript bronze were also
found in the tomb fill.
More important objects were also recovered from the tomb fill. At the
western part of the southern half of the chamber and close to the floor,
within the mud, were found 565 gold beads of varying types and 38 stone
beads.
The beads are all quite small, as may be seen in Figure 13. There were 431
flat, round beads (labeled 3), 3.5mm. in diameter and .5 mm. in height; 87
round beads with double carination (5), 5mm. in diameter and 1.5 mm. in
height; 40 hollow, spherical beads (6), 7mm. in diameter, with walls .5 mm.
thick; 4 very thin, flat, round beads (8), 5mm. in diameter and 5 mm. thick; 2
round, lentoid-shaped beads (4), 2.25mm. in diameter and .5 mm. thick; and
1 round, narrow-walled bead with a relatively large hole (not numbered),
4.5mm. in diameter and 3mm. thick.
Those of stone included 31 round carnelian beads with a slight double
carination (10), similar to but slightly larger than some of the gold examples,
6mm. in diameter and 2mm. in height; and 7 solid, round carnelian beads
(9), 7.5mm. in diameter and 4mm. in height. In addition there was one
simple flat bead apparently made of tortoise shell (11), 5 mm. in diameter
and 2.5mm. in height.
In the eastern part of the southern half of the chamber we found one flat
bronze adze and three bronze axe heads, all of the same type, but each made
in a separate mold (Figure 14). The bronzes were in excellent condition,
albeit they were found in the mud. The edges of all the blades were quite
sharp, and it would therefore seem that they belong to the original tomb
56 chapter two
Figure 13. Gold and stone beads from the tomb, Tumulus IV.
contents and were not the tools used by the robbers to dig into the tumulus.
Those tools were not left behind as they were needed to dig into the other
tumuli in the area!
The adze is 13.8cm. long and 3mm. thick; it flares out slightly from a width
of 3cm. at the base to 4.35cm. at the outer edge.
Each axe has a shaft hole close to the back part of the weapon, a single
oblique point forming the rear, and an outward-flaring blade. The three axes
have slightly different measurements: (12) length 14.3 cm., width 4.6 cm.;
(13) length 14.5cm., width 4.3cm.; (14) length 13.9 cm., width 4.5 cm.
The north-south trench was excavated for a length of 10 meters, measur-
ing south from the tomb edge, down to the level of the top of the tomb. At
a distance of 4.5 meters south of the inner edge of the tomb we cleared an
irregular section of rubble stones, three stones and 90 cm. wide, extending
east-west across the trench and disappearing into both undug balks (Fig-
ures 8, 10); 1.75 meters north of these stones was a single stone sticking out
of the west balk.
There can be little doubt that the section of stones represents part of a
ring that encircles the tomb, rather than the remains of a central rubble
Figure 14. Bronze axes and adze from the tomb, Tumulus IV.
58 chapter two
mass familiar to us from the other tumuli. No other stonesexcept the

unexplained odd one in the west balkare to be seen in any of the exposed
sections, either around the tomb or in the north-south trench. The sections
mentioned show clay not disturbed by the robbers and would show rubble
stones in situ if they had ever been laid down. One problem cannot be
resolved: where did the stones that were found on top of the tumulus next to
the robbers shaft come from if not from the area over the tomb itself? And
what relationship, if any, exists between these stones, the stone circle, and
those stones found within the tomb? It is possible that the area of the tomb
chamberbut not the tops of the walls of the tombalone was covered
with stones (what the roof consisted of is of course not known) and that
the robbers removed most or all of these stones through their meandering
tunnel and vertical shaft. But this suggestion cannot be proven and does
seem unlikely, so the issue will have to remain unresolved.
To the south of the ring of stones the fill consists of gravel and sandy
soil that form a bulge (Figure 6). To its north there is a layer of soft brown
earth under a thick layer of clay. It would appear that after the tomb pit had
been dug and the stone tomb constructed the area immediately adjacent
was leveled up to the stone ring, and that the bulge may represent dumping
during the digging of the pit and the leveling process. Directly over the
tomb (whether or not it was covered with stones) and the surrounding area,
grayish yellow clay in compact condition was laid down by dumping; on
top of this was dumped a mixture of clay and gravel. One and possibly two
cleavage lines, similar to those from Tumulus I (infra), were recognized,
and this fact suggests that the tumulus was erected with the aid of portable
partitions that held the clay while it was being laid down. Shortly before the
tumulus reached its final shape, a revetment of small stones in one or two
layers was placed around the lower part of the slope. The revetment was
irregular in height and was not compact, suggesting that it was laid down
in a hurry. Test trenches dug around the tumulus confirmed the presence of
the revetment around the whole perimeter (Figure 3). After this stage, gravel
and clay were dumped and the desired tumulus shape was formed.7
7 It has of course occurred to me that the revetments at S Girdan may actually have been
originally exposed and not covered with earth as they now are: that is, they are covered now
by earth from the upper part of the tumulus. However, the upper borders of the revetments
are never uniform, and there is no regularity in the manner in which they are laid down: gaps
and depressions, and shifts in levels, occur on all tumuli, as may be seen by looking at the
sections. One might conclude that early stone robbing would account for these irregularities.
I prefer to leave the matter open but suggest that the revetments were meant to be covered,
as concluded in the text.
Figure 15. Plan of Tumulus V.
Tumulus V
This tumulus lies about 100 meters to the northwest of Tumulus IV, in a
row with Tumuli I, II, and IV (S Girdan I, fig. 2, F on plan). Its present
height is about 5 meters, its diameter 48 to 50 meters. At present it is
assymetrical in shape with a deep pit at the top (Figure 15). There were
no clear indications as to the location of the precise high point of the
tumulus, so we arbitrarily chose the center of the pit as our center point
(infra); we assumed that whoever dug the pit picked the highest point as
the center. After the usual division of the tumulus into quadrants we chose
the upper part of the southwest quadrant for excavation, using true north as
orientation.
60 chapter two
Figure 16. Section AB, Tumulus V.
The fill consisted of gravel and clay; about 20 cm. below the surface we
began to encounter scattered stones. They covered a good part of the south-
ern area of the excavation but presented no pattern. Stones continued to be
found throughout the fill (in the southern area). The northern part of the
excavation, on the other hand, consisted of hard clay. After a time it became
clear that the softer gravelly clay mixed with the stones represented a dis-
turbed area, and we could see the faint outlines of an irregular pit (Figures 16,
17); the pit penetrated to a point just above the tomb subsequently discov-
ered.
In the northern part of the excavated area, the part consisting of hard clay,
five distinct cleavages were recognized in the section, and we were able to
isolate portions of them on the horizontal surface (Figure 18); other cracks in
the section may be cleavage lines or cracks from the sun, but we could not
tell. The five cleavage lines mentioned here are distinct and unmistakable
Figure 17. Section BC, Tumulus V.

62 chapter two
Figure 18. Cleavages, Tumulus V; AB section at right.
and, as with Tumuli I and IV, reflect the use of portable partitions.8 No
cleavages were recognized in the north-south section.
At a depth of about 4 meters large rubble stones mixed helter-skelter with
flat stones appeared in the west, north, and south areas of the excavation.
Unfortunately these stones turned out to be the remains of three sides of the
tomb (we did not excavate the fourth side) and the disturbed rubble overlay
(Figure 19).
The tomb was apparently rectangular in shape, about 2.25 meters in
width, and formed from a pit dug into the earth. It was oriented roughly
east-west, with the southern wall entering the BC section, the northern
wall entering the AB section. The sides were the earth walls of the pit
itself, but because of the havoc we could not tell if they had been plastered
or smoothed. The upper edges of the pit apparently had been lined with
irregular flat stones or slabs. We were able to surmise this information
because some slabs were found on the edge of the pit and also because of the
analogy with Tumulus VI (infra). Within the tomb pit some more slabs, also
irregular in shape, were found (Figure 19), but we are not able to conclude
8 The distances between cleavages were 17, 30, 12, and 15cm.
Figure 19. Tomb, Tumulus V; note the skull on the

floor and bones on the stone in the left foreground.
whether they represent a floor that was torn up or fallen slabs that originally
lined the edge of the tomb (cf. Tumulus VI). Otherwise, no floor could be
recognized; the deeper we excavated, the muddier the earth turned.
Soft grayish white ashy deposits were found mixed with the stones and
perhaps are the remains of a wood or reed roof, but we cannot be certain.
64 chapter two
The tomb had been ruthlessly torn apart by the robbers, making it impos-
sible to draw a plan. Originally a rubble-stone overlay covered the tomb, but
since this had been torn away, we found the rubble stones jumbled together
with the flat stones. Within the tomb were found fragments of bone scat-
tered about and part of the skull of a young adult male (Figure 19); a long
bone was found on top of some stones outside of the tomb at the north-
west edge. The only other objects recovered were a small gold bead, flat
and like one of the four gold beads found in Tumulus IV (Figure 13, no. 8); a
small, carinated, black and white stone bead, 1.2 cm. in diameter and 7 mm.
in height; and small, nondiagnostic fragments of bronze. All were found in
the disturbed fill around the tomb. Some pottery sherds were also found in
the tumulus fill. They are red-buff wares and generally non diagnostic (Fig-
ure 29) except for one sherd that was once part of a carinated bowl similar
to the one found in the stones on top of Tumulus IV (Figure 28).
Whether or not the tomb lies away from the center of the tumulus cannot
be stated because of the disturbance caused by the large pit. Since the high
point of the tumulus is now missing and the adjacent areas corrupted, we
have no objective guide. I will return to this matter shortly.
Test trenches were dug in the north and west quadrants, and a long trench
was dug from the main cut (Figures 15, 17, 20). These trenches revealed the
expected revetment of small stones that encircled the lower slopes of the
tumulus.
At the upper border of the revetment stones revealed in the test trench
in the western quadrant (Figure 15, X on plan) and just below the surface,
we found a redbuff-colored jar with an everted lip and a raised ridge at the
shoulder (Figure 31). Within the jar, which was in fragments, were found
badly crushed human bones, apparently those of an infant. The jar was
placed at the edge of the stones just touching them, implying perhaps that
those who deposited the burial knew about the revetment. Yet we cannot
rule out the possibility that the deposition at this particular place was acci-
dental. The vessel could be called an Iron II or III vessel, but I am reluctant
to make a more definite decision on the basis of a coarse, undecorated jar.
No other burials (except for an Islamic burial in Tumulus VI) were found
within the fill of the tumuli at S Girdan, but since we have not cleared
away all the upper fill of the tumuli, we are not in a position to make defini-
tive statements on the matter. In any event, even if the jar was buried at
the time of the erection of the tumulus, we do not know if the burial was a
significant event or simply an instance of someone taking advanage of the
tumulus as a convenience. I can see no reason to bring in a discussion of
sacrifice.
Figure 20. North-south trench with revetment stones, Tumulus V.

66 chapter two
Sir Aurel Stein mentioned that in his excavations of the tumuli in 1936
shafts were sunk on the top of a couple of these mounds 9 Stein did
not mention which tumuli he tested with shafts nor how deep his shafts
penetrated. We therefore do not know if the pit recognized in Tumulus V
is Steins work, although this is quite possible. For what seems fairly certain
to me is that the pit does not represent the work of those who plundered
the tomb: the pit does not penetrate as far as the tomb (Figures 16, 17). In
fact, it seems very probable that the robbing and destruction of the tomb
occurred before the erection of the tumulus. Hard clay exists directly over the
tomb, and the cleavages, surely representing a technique of construction,
were in situ in the fill over part of the destroyed area, the area not disturbed
by the later pit. And directly below the hard clay and the cleavages was
found the destroyed tomb. The only conclusion possible, it seems to me, is
that the tomb had been plundered and destroyed after the interment and
deposition of the grave goods, and that the mourners of the dead person
decided to erect the tumulus nevertheless. Perhaps we may assume that
the tomb was robbed as a result of an enemy or bandit raid. Following this
act of desecration the survivors decided not to dishonor the dead man by
leaving him unburied, and erected the tumulus; why they did not arrange
his scattered bones eludes us.10
An interesting parallel (archaeological, not historical) for the erection of
a tumulus after its tomb was plundered apparently exists in Tuekta, about
120km. west of Pazyryk, in the Altai region of eastern Russia. S.I. Rudenko
excavated two stone kurgans, dated to the late sixth century bc, neither
of which exhibited any signs of disturbance. Yet when the tombs were
reached and cleared, it became obvious that they had been robbed. The
conclusion seems to be that they were plundered before the tumulus was
erected.11
9Old Routes of Western Iran (London, 1940) p. 377.

10 E. Lorenz, Raubgrberei-nicht Aktenkundig, Antike Welt I (1970) pp. 39f., suggests that
graves and tombs were destroyed not only as a means of securing the contents, but as a
political and religious action against the entombed and his culture.
11 S.I. Rudenko, Kultura Naseleniia Tsentralnogo Altai u Skifskoe Vremia (Moscow, 1960)
pp. 93 ff., pls. XIII, XIV (I wish to thank Prof. Ann Farkas for helping me with translation); K.
jettmar, Art of the Steppes (New York, 1964) pp. 12 ff., figs. 104, 106.
Figure 21. Plan of Tumulus VI.
Tumulus VI
This tumulus is the last one in the row of seven counting southeast to
northwest (S Girdan I, fig. 2, G on plan). It is a relatively small and low
mound, with a preserved height of 2.5 meters and a diameter of about 30
to 38 meters (Figure 21). Like the other tumuli it also is surrounded by
cultivated fields.
In the top part of Tumulus G there was a large depression. Although
the depression could have resulted from Steins work, I was certain it was
evidence of plundering. Therefore I wished to excavate Tumulus H, less than
190 meters to the northeast of Tumulus IV. But because there was confusion
on the part of the local authority about whether or not H was part of the S
Girdan cemetery, I reluctantly had to excavate G.
The tumulus was divided into quadrants, and we excavated most of the
southwest quadrant and parts of others while clearing the tomb. It was
68 chapter two
Figure 22. View of the tomb with surrounding rubble mass,

and revetment stones in the foreground, Tumulus VI.
not possible to be sure about the location of the original high point of the
tumulus because of the disturbed nature of the area. We arbitrarily chose
the center of the depression as the center point of our quadrants.
Surface features, aside from the depression, consisted of many stones 30
to 50cm. in diameter lying around the lower edges of the tumulus. They
appeared to have been loosened from the revetment stones encircling the
tumulus. The upper part of the rubble revetment was exposed for the whole
length of the southwest quadrant, and the complete length of the revetment
was exposed in a narrow test trench in the northeast quadrant (Figures 21
23). The stones are of mixed sizes, 5 to 20cm. in diameter, laid down in two
or three courses. However, in the western half of the southwest quadrant the
upper part of the revetment consisted of one or two courses of rather large
slabs, similar in type to those lining the upper edge of the tomb.
The tomb is a pit cut into the earth and is an irregular oval in plan. It
is oriented northwest by southeast with an interior measurement of 4.20
by 2.25/50 meters. It had (as surmised) been plundered in the past, and a
section at its northwest end was destroyed. Because of the plundering and
Figure 23. Plan of the tomb and revetment stones,

Tumulus VI. Datum point at top of tumulus.
accompanying destruction it was not possible to discern if the walls had

been plastered or smoothed, or if the floor had been covered with slabs.
Several slabs were found in disorder lying flat and standing upright within
the tomb, but we could not establish if they represented floor slabs or if they
had fallen in from outside (Figure 24). Water began to seep into the pit at a
depth of 1.37 meters, and even if there had been a smoothed floor, we could
not have recognized it. (Our workmen were convinced we had excavated a
fountain and pool.) Not a single object or bone was found in the tomb, all
having been taken or destroyed by the robbers.
The upper perimeter of the tomb pit was lined with two to four courses
of stone slabs about 10cm. thick and in sections overlapping each other
(Figures 24, 25). These slabs were bordered by a large rubble mass, consisting
70 chapter two
Figure 24. Tomb of Tumulus VI with slabs being excavated.

Figure 25. Tomb of Tumulus VI with surrounding rubble mass.
partly of flat slabs but mostly of large and small stones, that formed a
rough circle around the tomb. No evidence in any of the sections exposed
suggested that the rubble stones ever extended over the tomb. Needless to
say, we do not know the extent of the area cleared by the robbers, and the
sections yielded no evidence in this matter, but it is doubtful to assume that
they cleared away the stones neatly and uniformly down to the level of the
tomb on all its sides. Therefore, it would seem that the rubble mass was
laid down around the tomb and never functioned as an overlay, otherwise
so common at S Girdan, as seen in Tumuli I, II, III, and V. In this respect,
Tumulus VI reminds us of the fact that the tomb within Tumulus IV (another
plundered and disturbed tomb) also seems not to have had a rubble overlay.
The upper part of the tumulus fill, judging from the section in the north-
south trench (Figure 26), consisted of light-colored clay and gravel. Below
this was a layer of compact gray clay with pebbles that partly overlay the
rubble stones, and next was a layer of compact light gray clay that was
72 chapter two
Figure 26. North-south section of the southwest quadrant, Tumulus VI.
packed firmly against the outer border of the rubble mass. Below this were
still another layer of clay, tan in color, and then dark and moist earth that
must be virgin soil. We observed no cleavages in the sections or in the
surfaces excavated.
A burial of an adult male was found about 38 cm. below the surface in a
section that partly overlapped the west wall of the tomb. The skeleton was
lying on its side and faced southwest; there were no objects with the burial,
but two stones had been placed about 20 cm. above the head. This burial
dates from Islamic times and has nothing to do with the tumulus and its
construction.
The Center of a Tumulus
An assumption has been made both in S Girdan I and in this report that the
present high point of the tumulus is its center as understood by the ancient
architects. Most archaeologists who have excavated tumuli and discussed
the tomb position seem to have taken this viewpoint without specifically
defining their terminology. Yet it is important to realize that we do not know
what shape a particular tumulus had in its original state, given more than
two millennia of wind and rain erosion, not to mention human activities.
Nor do we know if the tumulus was originally built so that the shape was
uniform in its dimensions, i.e., whether it had uniform contour lines on all
sides, or whether this effect was not required. And we do not really know if
the highest point of the tumulus was understood to function as the center,
and that this point was kept in mind after a tomb was built. Another item to
be remembered is that the original edge of the tumulus will always be buried
under the present level of the surrounding fields.12 And it seems probable
12 See S Girdan I, pp. 8 ff., figs. 6, 7.

to assume that this burying did not occur uniformly on all sides, so that the
plan of the tumulus will have been altered. In other words, the original shape
and geographical center of the tumulus may actually elude us.
The excavator of the tumulus at Takht-i-Suleiman in Iran stated that its
original center, and high point, had moved about 3 meters to the north-
northwest and did not correspond to the present high point (Spitze des
Hgels).13 I am hesitant to either reject or accept this conclusion because
to my mind it is apparently possible from reading the published section to
conclude that in fact the original and present high point are the same. This
would mean that the stone pile and wooden marker excavated there, not
under the present high point, were meant only to be a guide for the builders
up to a certain stage of the construction and were not meant to mark the
final high point of the completed tumulus.14 However, this conclusion is not
based on direct observation of the excavated section. In any event, no tomb
was found either under the stone marker or under the present high point,
thereby establishing that, whatever unit is used as a modern guide, the tomb
was in fact placed off-center.
Reexamining the evidence of the early excavations at Gordion, we find
that the Koertes used such terms as Gipfel, Mitte, Mittelpunkt, and Zen-
trum when discussing the geography of the tumulus.15 I assume that Gipfel
must signify the present high point. How they arrive at the term Mit-
telpunkt is not discussed, but they do state that the tombs within Tumuli II,
III, and IV were not under the Mittelpunkt but under the Gipfel, that is,
under the high point. At the same time, the grave of Tumulus I was almost
exactly under the Mittelpunkt, and that of Tumulus V was three meters
from the Mittelpunkt; the Gipfel is not mentioned.16 However, the later
13 H. Wiegartz, Die Ausgrabung am Tumulus (Tepe Madjid), Archologischer Anzeiger
1965, cols. 788 ff., especially 789. Unfortunately I did not become aware of this important
article until S Girdan I was published. W. Kleiss sent me an offprint of his and R.M. Boehmers
contribution (but not Wiegartzs) on the excavations at Takht-i-Suleiman, Die Grabungen
auf dem Zendan-i-Suleiman, Archologischer Anzeiger 1965, pp. 759ff., and I assumed it was
the only report: see S Girdan I, p. 4, note 22.
14 Wiegartz, Die Ausgrabung, cols. 795 f., Abb. 79, where the fill described as Kies-
Bnder would be the final course of earth laid down over the regular bands of Lehm, Kies,
and Brauner Lss. The Humus would be accumulated fill resulting from erosion and would
not be part of the original tumulus. Note that my conclusion would better fit the suggestion
that the stone circle surrounding the tumulus was originally exposed; see Wiegartz, Die
Ausgrabung, col. 792.
15 G. and A. Koerte, Gordion Ergebnisse der Ausgrabung im Jahre 1900 ( Jahrbuch des Deut-
schen Archologischen Institut, Ergnzungsheft V; Berlin, 1904) pp. 38f., 99, 104f., 129f., 139f.
16 Koerte, Gordion, pp. 129, 139. My comments in S Girdan I, p. 22, note 11, about the tomb
of Tumulus III being under the center should be corrected to say under the Gipfel. The
74 chapter two
excavator at Gordion states quite definitely that the tombs he excavated

were not under the center (i.e., high point, or peak, to use his term) of the
tumuli, but off-center (and, moreover, in the southwest quadrant).17
The excavators at Sardis inform us that some of the tumuli are not under
the present high point, while others are. And the excavator at Kerkenes Dagh
claimed that by digging into the centers he could not locate the tombs in
some of the tumuli he excavated.18
Within Tumulus 3 on Cyprus the excavator found a brick beehive-shaped
structure, the center of which he interpreted as being the center and high
point of the tumulus (the high point is now gone). Because the tomb cham-
ber did not correspond to the position of the center of the brick structure,
he concluded that the tomb was off-center, i.e., not under the high point of
the tumulus. We are also told that at the same cemetery the tomb, actually
a cenotaph, within Tumulus 77 ne se trouve pas au centre du tumulus 19
In Europe, where the tombs appear always to be under the center of the
tumulus, it is the present high point that is used as a guide.
In short, observations about a tomb placed off or under the center of a
tumulus are usually (but see the Koertes at Gordion) based on the position
the tomb has relative to the present high point. It is not easy to decide if
this method is actually the correct way to judge if a tomb was consciously
and originally placed away from or placed under the tumulus high point.
Nevertheless, my own conclusion is that the ancient builders considered the
top of the tumulus to be the point of orientation, the center, so to speak,
whether or not it was in fact geographically so. Therefore, I do not wish
to alter my opinions about the off-center placement of some tombs within
confusion is of course the reason for this present discussion on terminology. Note also that
both the Koertes and R.S. Young found wood masts or markers over some of the tombs at
Gordion: the Koertes found them under the Gipfel, the latter over the tomb. Did they mark
off the high point and the tomb, or only the tomb? Following Young, they presumably marked
off not the peak but the tomb. See p. 22 and note 8 of S Girdan I.
17 The references are in S Girdan I, p. 22, note 11. In the same footnote appear two different
words used by T. zgu and M. Akok for describing the position of tombs within Tumuli I
and 2 at Ankara: under the Gipfel for I, under the Mitte for 2. In 1969 a Phrygian tumulus
was excavated near Ankara by archaeologists from the Middle East Technical University.
The tomb seems to have been placed off-center, away from the present high point; it is still
unpublished. Note that the Koertes, Gordion, p. 129, refer to the tomb of Tumulus I being in
the southwestern quadrant of the tumulus fill, using the Kuppe des Hgels as the center.
18 S Girdan I, pp. 22 f., notes 11, 13.
19 S Girdan I, p. 23, note 14; now also see V. Karageorghis, Excavations in the Necropolis of
Salamis (Nicosia, 1967) pp. 25 ff., 121 f.; V.`Karageorghis, Salamis: Recent Discoveries in Cyprus
(New York, 1969) pp. 71 f., 151 ff.
tumuli in Anatolia, in Cyprus, and at S Girdan in Iran;20 I only wish to

explain the criterion for such statements. But if it can be demonstrated (by
a geologist?) that the present high points of tumuli are arbitrary, resulting
from erosion and thereby creating a new configuration in the shape of the
tumulus (as stated by the excavators at Takht-i-Suleiman), we may have to
abandon any assignment of significance to tomb placement.
Conclusions and Summary
The three tumuli excavated this season share general features with the two
excavated in 1968: Tumulus IV and probably also Tumulus I (unexcavated)
contained tombs built off-center (we can say nothing definite about the
tombs in the disturbed Tumuli V and VI); all the tombs were built into
pits cut into the earth; all the tumuli have encircling stone revetments;
Tumulus V had a rubble-stone overlay covering the tomb; Tumuli IV and
V contained cleavage lines; and the stone tomb in Tumulus IV is of the same
type as that in Tumulus II.
Within this area of agreement, differences do occur, demonstrating that
variety did exist and that no rigid system of tomb architecture obtained.
Tumuli IV and V contained pit tombs, the top borders of which were lined
with slabs of stone. The plain pit tomb excavated in 1968 in Tumulus III did
not have a stone-lined border. The tomb ofTumulus IV apparently did not
have a rubble-stone overlay, but rather it had a feature unique in the S
Girdan series, namely, a narrow rubble wall that must have encircled the
tomb. And Tumulus VI apparently also did not have a true rubble overlay
covering the tomb but a variety in the form of a packing laid down around
the tomb. Finally, Tumulus VI is also unique at S Girdan for its roughly
ovalshaped tomb pit; the tomb plans of all the other tumuli are rectangular.
Some general comments about tumuli as well as foreign parallels for the
S Girdan tombs and tumuli have been presented in the first report; a few
comments will be added here, although I make no claim that all sources
have been covered.
20 I originally thought that by using the upper border of the stone revetment as a circle in
each tumulus, I could arrive at a true center point. I subsequently realized that this method
would not work as there was no regularity in the position of these stones around the tumulus,
and therefore I could not get a true circle; see, for example, S Girdan I, p. 9, fig. 5. For a brief
discussion of the possible relationship of Lydian and Phrygian tumuli (and Cypriote tumuli
also), see my article Phrygian or Lydian? Journal of Near Eastern Studies 30 (1971) p. 63.
76 chapter two
Within Iran itself one must refer to the two tumuli at Takht-i-Suleiman,
one of which, Tepe Majid, has been partly excavated (supra). This tumulus is
larger than any at S Girdan. Aside from the conical rubble pile and wooden
marker mentioned previously, a circle of stones, 1.50 meters wide, extended
around the base of the tumulus. According to the excavator this circle was
originally exposed. This is a feature shared with some tumuli in Europe and
the Caucasus.21 The technique of tumulus construction was not the same as
that recognized at S Girdan: at the latter site there were no central rubble
piles with wooden masts and no outer circle of stones, and the earth was not
laid down in the uniform manner observed at Tepe Majid. The significance
of this will have to await the excavation of the tomb that no doubt lies within
the tumulus. It has already been mentioned that no tomb was found at the
center of the tumulus.
W. Kleiss recently published a plan of a tumulus from the Ardebil region
west of the Caspian Sea.22 The tomb was constructed of stone and built into
the center of the tumulus; it was oval in plan, reminding us of the plan of the
tomb in Tumulus VI. No rubble-stone overlay covered the tomb, but there
was a stone circle around the perimeter of the tumulus.
Another tumulus in Iran on which Kleiss reported lies at the foot of
the Iranian-Urartian site of Bustam, 35km. north of Khoy; it is still to be
excavated and we have no data on it.23 One wonders if there can be any
significance in the fact that the cemetery at S Girdan also lies close to an
Urartian site, Qalatgah.
At Bogazky in Anatolia a tomb was excavated in 1958 that may have been
originally placed under a tumulus, although this is not certain because of
disturbances in the area.24 The tomb is brought into discussion here because,
21 References in S Girdan I, pp. 22 f., notes 15, 18. N.G.L. Hammond, Tumulus-burial in
Albania, the Grave Circles of Mycenae, and the Indo-Europeans, The Annual of the British
School at Athens 62 (1967) pp. 77 ff., has written about Albanian and Greek Bronze Age tumuli.
He discusses the House of Tiles at Lerna (see S Girdan I, p. 23, note 15) and grave circle B
at Mycenae as tumuli with encircling stones. See also N. Yalouris, A Mycenaean Tumulus
at Samikos, Deltion 20 (1965) pp. 6 ff. (Greek with French resum), a tumulus with a stone
circle; M. Ervin, News Letter from Greece, American Journal of Archaeology 74 (1970) p. 264,
for Bronze Age tumuli (Early Helladic date?) at Marathon.
22 Kleiss, Bericht ber zwei Erkundungsfahrten, p. 19, fig. 16; S Girdan I, p. 24.
23 Kleiss, Urartische Pltze, p. 23. The conical mound at Tusikarn on the road from
Kangavar to Jowkar looks to me as though it may be a tumulus, but it has not yet been
excavated. There is a puzzling reference to a tumulus burial in Persia where Clearchus
perished: Plutarch Artaxerxes 18. 5. Does Plutarch record an actual tumulus burial?
24 Peter Neve, Die Grabungen im Wohnviertel J-K/20, Mitteilungen der Deutschen Orient-
Gesellschaft zu Berlin 91 (1958) pp. 3 ff., Abb. 1a, 1b.

to judge from the published plan, it was surrounded by a circular rubble-

stone mass. This rubble mass did not cover, i.e., overlay, the tomb but was
laid down against and around it, in the same manner we observed to occur
with the tomb in Tumulus VI (Figures 22, 24, 25). The Bogazky tomb could
not be dated by objects, as none were found, but the excavator suggested it
was built in Hellenistic times.
Two tumuli have been reported in Syria by a German survey team.25 Each
is surrounded by a stone circle, and at one point the stone circles touch
each other. These tumuli have not yet been excavated. Other tumulus-like
mounds have been reported in an area near Jerusalem, but no graves have
been found within them.26 Since the remains of platforms and steps are
present, it has been concluded that the tumuli (or mounds) are the remains
of ritual areas rather than coverings for burials.
At least three tumuli excavated at Trialeti in Georgia appear to contain
tombs built away from the center. It is not clear to me if other tumuli at
Trialeti also have this feature, as the texts relating to the excavations do not
mention the tomb position; the evidence comes only from an examination
of the plans and sections.27 Kurgan IV and X, dated by B.A. Kuftin to the Early
Bronze period, and Kurgan V, dated by O.M. Japaridzi to the Middle Bronze
period, have pit tombs in an area definitely away from the high point of
the tumulus. These examples from Trialeti are the earliest examples of this
feature known to me.
The finds from the three tumuli excavated at S Girdan in 1970 were few, as
we have seen. Fortunately, several of the sherds found in the fill of Tumuli V
and VI furnish us with some information about chronology. Several of the
25 E. Heinrich, Die Vorbereitung: Bericht ber die im Euphrattal bei Aleppo begonne-
nen archologischen Untersuchungen, Mitteilungen der Deutschen Orient Gesellschaft zu

Berlin 101 (1969) pp. 33, 35, Abb. 6. It is not certain in my mind that these are actually
tumuli. A. Westenholz, Bertum, dam tum, and Old Akkadian Ki Gal: Burial of Dead Enemies
in Ancient Mesopotamia, Archiv fr Orientforschung 23 (1970) pp. 27ff., discusses textual
references to what appear to be tumuli to cover mass burials.
26 Ruth Amiran, The Tumuli West of Jerusalem, Israel Exploration Journal 8 (1958) pp.
205ff. I should also like to call attention to the fact that the tumuli excavated on Bahrein
Island have their tombs placed under the center, and they also have a surrounding stone
wall: E. Mackay et al., Bahrein and Hemamieh (London, 1929) pp. 3ff., pl. iv; G. Bibby, Looking
for Dilmun (New York, 1969) pp. 59 ff., pl. ii.
27 B.A. Kuftin, Archaeological Excavations in Trialeti, I (Tiflis, 1941) (in Russian) pp. 101ff.,
figs. 107, 108; Claude Schaeffer, Stratigraphie compare et chronologie de lAsie occidentale
(London, 1948) p. 506, fig. 41; O.M. Japaridzi, Archaeological Excavations in Trialeti (Tiflis,
1969) p. 76, fig. 47; l owe this last reference to Ann Farkas.
78 chapter two
Figure 27. Gray ware bowl, smoke blackened, from

stones below the surface at the top of Tumulus IV.
sherds come from deep bowls with plain incurving or outcurving sides
(Figures 29, 30), and one sherd from Tumulus V (Figure 28) has an incurved
rim and concave sides, representing a shallow bowl. Parallels for the vessels
represented by the sherds occur in levels of the Iron III period at several sites
in Iran, viz., Baba Jan, Godin, Hasanlu, Nush-i-jan, Zendan, and Ziwiye.28 On
the assumption that the sherds in the tumulus fill represent either earlier
or contemporary material that was inadvertently dumped as fill, we have
a terminus post quem date of Iron III for the tumuli. The vessel used as a
container for the childs bones found in Tumulus V (Figure 31) seems to fit
into an Iron II or III background, but I am reluctant to state this in absolute
terms.29
28 Bowls: C. Goff Meade, Luristan in the First Half of the First Millennium bc, Iran
6 (1968) p. 122, fig. 10, nos. 1, 12; T. Cuyler Young, Jr., A Comparative Ceramic Chronol-
ogy for Western Iran, 1500500 bc, Iran 3 (1965) p. 56, fig. 2, no. 6, p. 54, fig. 1, nos. 1, 2, 4,
p. 58, fig. 3, nos. 1, 6, 11, 17; R.M. Boehmer, Forschungen am Zendan-i-Suleiman in Persisch-
Azerbeidschan 19581964, Archologischer Anzeiger 1967, p. 580, fig. 8; D. Stronach, Excava-
tions at Tepe Nush-i-Jan, 1967 Iran 7 (1969) p. 17, fig. 6, nos. 13. Bowls with incurved rim and
concave sides: Goff Meade, Luristan, p. 122, fig. 10, no. 3; T. Cuyler Young, Jr., A Comparative
Ceramic Chronology, p. 54, fig. 1, no. 6, p. 56, fig. 2, no. 10, p. 60, fig. 4, nos. 7, 12; T. Cuyler Young,
Jr., Excavations at Godin Tepe: First Progress Report (Toronto, 1969) p. 119, fig. 43, nos. 4, 5, 10,
p. 123, fig. 44, nos. 6, 7, 9, 11, 14, 15, 17; Kleiss and Boehmer, Die Grabungen, pp. 759f., fig. 72,
nos. 4, 5, 6; see also some close parallels in R. Ghirshman, Village Perse-Achmnide, Mmoires
de la Mission Archologique en Iran 36 (Paris, 1954) pl. xxxvii, nos. G.S. 1219d, G.S. 1224 from
level 2; there are also good examples from Agrab Tepe and Pasargadae, not yet published.
29 See, for example, T. Cuyler Young, Jr., A Comparative Ceramic Chronology, p. 56, fig. 2,
no. 9 (with handles); p. 63, fig. 6, nos. 1, 9; p. 65, fig. 7, no. 9.

Figure 28. Buff ware, slipped, sherd from

the fill of Tumulus V. Diameter 12cm.
Figure 29. Sherds from the fill of Tumulus V.
Figure 30. Sherds from the fill of Tumulus VI.

80 chapter two
Figure 31. Red buff jar from Tumulus V. Height: 32cm.
The nearly complete bowl found at the top of Tumulus VI (Figure 27)
within the stone debris close to the robbers shaft is indeed a good Iron III
vessel similar to the sherd from Tumulus V mentioned above (Figure 28).30
Its presence near the robbers shaft and its broken state surely indicate that
it is associated in some manner with the robbers activity. But did it come
from the tomb itself as booty, then to be dropped and abandoned? Or was it
the personal bowl of one of the robbers, brought with him to hold his yogurt?
We do not know, of course; but at least we have a terminus ante quem date
for the tumulus, also Iron III or earlier. In this respect we have been able to
reinforce the suggested dating for the S Girdan cemetery proposed in the
first report (S Girdan I, p. 24).
The axes (Figure 14) present a more difficult problem in terms of chronol-
ogy and foreign parallels because I cannot find any other axes of exactly the
same shape with the single sloping rear point. Axes with flaring blades and
multiple rear points are quite common in the Near East from very early times
continuing into the first millennium bc.31 At present it seems to me that it
would be correct to date the blades tentatively to the seventh or sixth cen-
tury bc on the basis of the archaeological interpretation reached for the date
of the tumuli.
30 See note 27 for parallels.

31 Jean Deshayes, Les Outils de Bronze de lIndus au Danube, II (Paris, 1960) pls. xviii ff.
The closest example I could find is an iron pick excavated by Layard in the North West
Palace at Nimrud, a structure built by Ashurnasirpal II and restored by Sargon II: A.H. Layard,
Discoveries Among the Ruins of Nineveh and Babylon (New York, 1875) p. 165, fig. at top.
The gold beads from Tumulus I (Figure 13) are very similar to those found
in the tomb of Tumulus III in 1968. However, they are not characteristic of
any one particular period, and we cannot discuss them chronologically.
No evidence exists that would allow us to decide which tumuli are earlier
and which later. Looking at the plan in S Girdan I, p. 6, fig. 2, we see that
seven tumuli exist in a row placed roughly east-west. They are all spread out
one from the other except for Tumulus II, which seems to have been tucked
in between Tumulus I and IV, implying perhaps that it was built after those
two were in existence. But we do not know which tumulus in the row was
built first.
Tumuli H, I, J, and K exist outside of the row and are spread out in no
apparent order. What their chronological relationship is to the others is of
course not known, and guessing will not help. We must, therefore, conclude
that the cemetery at S Girdan is an Iron III creation, perhaps seventh or
sixth century bc, and not make any finer distinctions.
chapter three
THE CHRONOLOGY AND CULTURE OF S GIRDAN: PHASE III*
In 1968 and 1970 six of eleven surveyed tumuli were excavated by the author
at a site called S Girdan, a few miles northeast of Dinkha Tepe and about
15 miles southwest of Hasanlu and Lake Urmia in northwestern Iran (Fig. 1);
the reports of the excavation were published in 1969 and 1971.1 The six tumuli
(labeled I to VI) were laid out in a roughly straight row, NWSE. Four of
their tombs had been plundered at some time in the past. All the tombs
were neatly cut pits; all were rectangular, except VI, which was an irregular
oval; II and IV had stone lined walls; III was a plain pit, while V and VI had
only the upper border of their pits lined with courses of flat stones; II and III
(IV?) had pebble floors; all the tombs but IV had a stone rubble overlay; and
all tumuli were encircled at their perimeters by a rubble stone revetment.
The tomb of Tumulus I (the largest in height, 8.25 m, Fig. 2) was not found,
and only Tumulus III and its tomb was recovered intact (Fig. 3). Relatively
few artifacts were recovered. From the plundered Tumulus IV (Fig. 4) were
retrieved three socketed axes and one adze (Fig. 5), and 565 small gold and
38 flat and rounded carnelian beads. The intact Tumulus III contained a
feline-headed stone scepter or whetstone, a fragmented silver vessel, an
adze, a knife, many stone and paste beads, and two fragmented silver rods.
A few nondescript sherds and vessels were recovered from the surface and
fill of Tumulus II, IV, V and VI; also, from the fill of Tumulus I, a circular lid
with a solid knob made of a reddish fabric with inclusions (Fig. 6). On the
basis of these artifacts, for which I could find few if any comparanda, and
depending primarily on tumuli comparisons in Anatolia, and elsewhere, I
dated the tumuli to the Iron III period of Iranian archaeology, the 7th6th
centuriesbc.2
* This article originally appeared as The Chronology and Culture of S Girdan: Phase III,
Ancient Civilizations from Scythia to Siberia 9, nos. 12 (2003): 117131.

1 Muscarella 1969, 525; idem 1971, 528.
2 For the artifact references see Muscarella 1969, 2021, figs. 2729; Muscarella 1971, 14,
figs. 13, 14; for the proposed chronology see Muscarella 1971, 2224; and Muscarella 1971, 2328.
For some unexplainable reason I neglected to mention the lid in Fig. 6 in these publications.
It may supply an ante quem non date for the tumulus.
84 chapter three
Figure 1. Map of Urmia Region.

the chronology and culture of s girdan: phase iii 85
Figure 2. View of Tumulus I, southeast from top of Tumulus IV.
Figure 3. Tomb of Tumulus III.

86 chapter three
Figure 4. Tomb of Tumulus III.
Figure 5. Axes and adzes from Tumulus IV.

Figure 6. Reddish fabric lid, fill of Tumulus I.
A second phase of discussion of the tumuli occurred in 1973, when in

the journal Iran the French archaeologist Jean Deshayes wrote a critique
of my chronology.3 It was far too late he argued, the tumuli were in fact
culturally and chronologically related to those of the Early Bronze Age of
southern Russia, the northern Caucasus. In particular, the axes and adzes
were similar in form to examples from this period in the northern Caucasus.
The same formal similarity with the Caucasus obtained for both the tumuli
and the tomb construction, where some tombs there had rubble walls; also
these tombs shared the presence of ochre on the bones; and there was an
absence of iron at S Girdan. To him, the latter sites artifacts and tomb
construction documented the appearance from the north of the earliest
Indo-European settlers in northwestern Iran in the last centuries of the 3rd
millennium bc, some 1500 years earlier than Iron III. Responding in the
same issue of Iran4 I disagreed with the significance, the exactness of his
comparisons and rejected his conclusions. We both agreed that the pottery
in the tumuli fill was not chronologically conclusive, although to him there
3 Deshayes 1973, 176178. Deshayes generously shared his article with me in advance of
its publication.
4 Muscarella 1973, 178180.
88 chapter three
were 3rd millennium parallels more available than my proposed later dated
examples, and which I continued to accept. I recognized no observable
relationships between the Caucasian axes,5 the adzes, or with the (plain!)
scepters he brought forth as comparanda. Nor did the nature of the tumuli
constructions he presented reveal relevant comparative information. They
did share the off-center from the tumulus peak tomb placement, but the
rare rubble walled tombs were not the same as the neat, well-cut slabs at
S Girdan; and rubble revetments surrounding each tumulusa distinctive
feature of the S Girdan tumuli (not mentioned by Deshayes; see below)
were absent; further, ochre continued in use over the millennia.
After re-reading the original reports and the 1973 articles I believe that
given what was presented as evidence by Deshayes for an earlier chronology,
my response was not archaeologically unreasoned. Perhaps in hindsight
more attention should have been paid to the absence of iron in the tombs
and relevance of the presence of ochre (for copper, see below). But at that
time, of course, Scythians and Cimmerians were on my mind, and I did not
have the evidence now availablesignificant data unknown to Deshayes
and me.
In May 2002, three decades after the last discussion, Elena Izbitser, a Rus-
sian scholar at the Metropolitan Museum of Art, brought to my attention
an article by B.A. Trifonov and a book by A.D. Rezepkin, whose works, pub-
lished in the same year (2000), challenged my late dating and argued that
the S Girdan tumuli were constructed in the Early Bronze period.6 Trifonov
called attention to the uniqueness of tumuli in Iran and the lack of com-
parative material available at the time of excavation, both, he recognized,
causing difficulties for recognition of a correct chronology. But now, how-
ever, new information demands that the sites chronology be reevaluated
and addressed anew. Both authors, Trifonov in more detail and analyses,
demonstrate that within the Maikop cultural period in the northern Cau-
casus tumuli at several sites share a crucial feature with those at S Girdan:
the outer encircling stone revetment (Fig. 7). As for the tombs, they share
5 He cited examples from his book (Deshayes 1960), one of which, no. 2103, was not exca-
vated and has little to do with the S Girdan examples: compare the axe parallels suggested
by Deshayes with those mentioned below and in note 8, which are indeed acceptable paral-
lels. Note that the closest parallel I could then find was an axe from Nimrud, Muscarella 1971,
25, footnote 31; Muscarella 1988, 81.
6 Trifonov 2000 (in Russiantranslated for me by Dr. Izbitser); Rezepkin 2000. Trifonov,
246, footnote 1, cites the scholar G.N. Kurochkin and an earlier (1989) article by Rezepkin
where this early chronology for S Girdan had previously been expressed.
Figure 7. Tumulus constructions. 1. S Girdan, Tumulus III;

2. Kishpek, kurgan 2. After Trifinov 2000, fig. 1.
a simple pit or stone lined walls, pebble floors, the tomb covered by rubble
stones, also body positions, and the presence of ochre on the bones.7
Further, and fundamental to the issues of chronology and culture under
review, both authors presented examples of socketed axes from several
Maikop period sites, Lechinkai and Pyatigorsk, that are obvious and rele-
vant parallels to the S Girdan examples, a bent butt and a blade with a
curved base (Fig. 8).8 Among these parallels is a hoard of socketed axes and
7 Trifonov 2000, 244, 255256, fig. 1 for a neat juxtaposition of the tumulus construction
of Tumulus III at S Girdan and one from the north Caucasus (here Fig. 7); Rezepkin 2000,
2021, 54, Taf. 10,11, 20, 26, 27, 32, 38, 49, 51, 63, 64, 74, 75, 76; Munchaev 1994, 184, fig. 45, 1.
For similar tumulus and comparable tomb constructions see Batchaev 1984, 116, 118, 120
fig.; Chechenov 1984, 185, 195, 207fig. Trifonov 2000, 256exaggerated the extent of the
agreements of the body and head orientation between the two regions under review. Note
that neither author was aware of the 1973 Deshayes and Muscarella articles.
8 Trifonov 2000, 257: compare his fig. 3 (here Fig. 8), nos. 13, axes from S Girdan with
those from the northern Caucasus, nos. 6, 8 (his adze comparison is not so close); no. 7 is from
Tianeti in the northern part of the southern Caucasus (see Kushnareva 1997, fig. 18, no. 23).
See also Rezepkin 2000, 6, 2021, Abb. 2, nos. 64, 65; Chernkh 1992, fig. 24: 24, 25. Rezepkin
2000, 20: cites the article by Batchaev and Korenevskii 1980, 7983 on the Lechinkai site that
shows the same relevant Maikop period axes noted by Trifonov, including an example from
Pyatigorsk in the Northern Caucasus, fig. 3: 3 (neither these authors, nor Chernkh, refer to
90 chapter three
Figure 8. Socketed axes. 14. S Girdan, Tumulus IV; 5. S Girdan,

Tumulus III; 6. The Yerevan hoard; 7. Tianeti; 8. Lechinkay;
9. The Maikop kurgan (Oshad). After Trifonov 2000, fig. 3.
adzes from a site near Erevan (called Prierivanian hoard in Russian), Arme-
nia, south of the homeland and half the distance between S Girdan and
the northern Caucasus.9 Trifinov also cited the silver vessel from Tumulus III
as fitting into a Maikop background (also A. Sagona, personal communica-
tion).10
In my 1971 publication I assumed that the S Girdan axes and adzes
were bronze, but in the 1980s an axe and adze from the site in the Metro-
politan Museum were tested and disclosed to be 98.6 and 99.3 % copper
the S Girdan or Deshayes publications). Trifonov correctly notes that the S Girdan adzes
are not identical to the Caucasian examples.
9 Trifinov 2000, 257; see also his satellite photo, p. 245, and Munchaev 1994, 171, fig. 3 for a
map of sites. The Erevan axes, not the adzes, closely parallel those from S Girdan. Chernkh
1992, 6465, fig. 21, had previously discussed these axes but he could find no parallels and
tentatively attributed them to the later KuraAraxes, Transcaucasian period.
10 Trifinov 2000, 258; see also Munchaev 1994, 201, fig. 51.
respectively, with minor trace elements.11 In September 2002 I requested a

new examination. Using ESD analysis, it was determined that the axe was
composed of 98 % copper and 1.6% arsenic, and the adze 99.2 % copper
and 0.7% arsenic; in both artifacts traces of nickel, zinc, tin and lead were
not detected. That they are composed of copper with arsenicare arsenical
copper, not bronzeis an important characteristic that marks their manu-
facture in the Early Bronze period.12
Based on the new evidence brought forth, it seems clear that S Girdan
is to be perceived and appropriately studied as a northwestern Iranian
manifestation of the North Caucasian, Maikop, Early Bronze Age culture;
it is not an Iron III site. J. Deshayes was essentially right after all, I was quite
wrong.
Of significance for future chronological refinements within the long
Maikop period is the fact that notwithstanding the cultural and general
chronological association of S Girdan and the North Caucasus group, a
number of differences are evident. At S Girdan the slab-built tombs are bet-
ter made, more painstakingly constructed, also larger (Fig. 4),13 compared to
the rubble walled tombs of those known in the Caucasus. One might expect
more examples than exist of the relevant socketed axes in the homeland
several occur further south in the Caucasus, and a number from several
Maikop are not close.14 Not a single example of an animal headed scepter
has yet been recovered in North Caucasus burialsor elsewhere.15 I know
of only one exact parallel, also with a feline head (but 43.5 cm in length,
longer by 6.5cm than the S Girdan example): it was plundered from some-
where to be sold and we can never know its archaeological provenience.16
11 Muscarella 1988, 81, nos. 143, 144, footnotes 1, 2; arsenic was not mentioned. The two
objects came to the Museum as part of the excavation division with Iran.
12 Trifonov uses the term bronze to describe not only these artifacts but also those from
the culture in general; Chernkh also does the same throughout his book. I suggest this is
incorrect and misleadingthe metal is copper or, arsenical copper, not arsenical bronze.
Lyonnet 2000, 202, correctly uses cuivre arsenic. See Chernkh 1992, 74 on arsenic as a
component of Maikop period copper artifacts.
13 See Muscarella 1969, figs. 1316; Muscarella 1971, figs. 5, 812, 24, 25.
14 Munchaev 1994, 192, fig. 47.
15 Burney and Lang 1971, 77 mention without references a pig/horse headed whetstone
from a tumulus near Lenkoran. Trifinov 2000, 258259 notes that there are no Maikop
examples, and he does not present relevant parallels. The animal headed Tierkopfzepter
shown in Rezepkin 2000, 14, Abb. 5 are not relevant.
16 Sotheby Parke Bernet Inc. New York, May 20, 1982, no. 29: the S Girdan example is
cited as a parallel. Indeed, it could have derived from one of the six unexcavated tumuli at S
Girdan from recent, post 1970 plunder.
92 chapter three
And there was no pottery recovered in the single intact tomb at S Gir-
dan (Tumulus III), and none recovered (abandoned) in the plundered ones,
which is certainly strange, an anomaly. Whether these differences collec-
tively reflect a cultural or, rather a chronological divergence between the
two distant regions warrant future research and discussion.
Tumuli are rare in Iran but two further examples are archaeologically
recorded in northwestern Iran. Although not fully excavated, they neverthe-
less preserve important evidence that allows them to be cited, if tentatively,
as culturally and chronologically relevant to our tumuli. Their existence
not too distant from each other or from S Girdan, suggests that the lat-
ter tumuli were not an isolated occurrence, although the latter alone may
be characterized as a planned necropolis. At Zendan-i Suleiman, north of
Takab, southeast of S Girdan, a partially excavated tumulus (the tomb was
never discovered) has an encircling revetment wall and the tomb is surely
off-center.17 The other tumulus, only surveyed, is 27 km west of Ardebil, west
of the Caspian, to the northeast of both S Girdan and Zendan; it too has
an outer revetment wall.18 This noteworthy feature characterizes the three
tumuli sites recorded to date in northwestern Iran.
This phase III of the investigation thus allows us to identify the S Gir-
dan (and perhaps the other two mentioned) tumuli as components of the
Early Bronze/Maikop cultural period. Their historical reality is all the more
complex and noteworthy when situated in the local historical background.
To date no Maikop period sites have been identified in northwestern Iran,
but in fact Maikop period material even in its homeland is archaeologically
known primarily from burials. However, a very large number of excavated
and surveyed sites, about 35, were identified by an Italian team in the plain
and foothills on the western shore of Lake Urmia as having Early Bronze
Age Transcaucasian (ETC) wares of its various sequential phases. The exca-
vated sites include Geoy Tepe (Period K), Haftavan, and Tappeh Giljar.19 On
17 Wiegartz 1965, 788799, Abb. 78; Muscarella 1971, 23, 25.

18 Kleiss 1969, 19, Abb. 16; Muscarella 1971, 25, where see also footnote 23 for mention of an
apparent tumulus (unexcavated) at Tusikarn, near Kangavar. Its contents and date can only
be determined by excavation.
19 Pecorella, Salvini et al. 1984, 244, 248, 332333, pl. XLIV, figs. 65, 6674; 142ff. are listed
the relevant ETC sites they discovered: numbers 2, 3, 7, 12, 16, 19, 2023, 33, 34, 36, 42, 51, 55, 57,
60, 65,7275, 7981, 87, 88, 97, 98, all placed on the map opposite on 179 (no. 16, Tappeh Giljar,
sometimes written Gizler, is a major long lasting ETC settlement). For other discussions of
ETC sites in NW Iran see Burney and Lang 1971, 45, 53, 5961; Dyson 1973, 697699; Sagona
1984, 6065, 241242, 246247; Kushnareva 1997, 43 ff., especially 65, 69, 70 (but missing here
the eastern side of the lake is the excavated site Yanik Tepe, and about 9
unpublished sites surveyed by W. Kleiss that yielded ETC wares.20 Some 45
or more ETC sites have also been identified to the south of Lake Urmia,
including Hasanlu Period VII, which has a couple of classic ETC sherds,
and also in an earlier context, pottery of the earliest ETC, phase, and Godin-
Tepe.21 They occur east of the Mahidasht, close to Malyer; the range for Iran
is ca. 30002200 bc.22 In general, then, they may all disagree more or less
with the incipient date, but not the terminal one.
New chronologies for the Maikop period from C 14 data suggest a. range
from ca. 3700/3500 to 3000/2900bc, a long time. Trifonov dates S Girdan to
the second half of the 4th millennium bc, ca. 35003200bc, i.e. presumably
within the cultures middle to late phases (and earlier than the initial ETC
dates suggested by Sagona and Kushnareva). And he suggests that rather
than being contemporary to the local ETC sites of Urmia, the tumuli are
more likely to have been constructed before their arrival.23 If his chronology
is essentially correct, and noting some of the incipient ETC chronologies
suggested, the S Girdan tumuli are consequently artifacts of an intrusive
cultural occupation in northwestern Iran at a time earlier than the ETC
incursion there; they represent its precursor. An alternative interpretation
might propose that the late Maikop and ETC cultures overlapped in part in
northwestern Iran, that both cultures co-existed side by side for some time.
This view surfaces because the proposed chronologies suggest that the ETC
are the important Pecorella, Salvini et al. and Sagona publications, any mention of Godin
Tepe; and map on p. 46 is useless on the NW Iranian ETC sitesnone are cited; the problem
exists also with Lyonnet 2000, fig. 1); Meyer 2001 also omits Pecorella, Salvini et al. 1984; on
314 he lists an incomplete 43 sites on the west and northern areasup to the Araxes.
20 Meyer 2001, 314, n. 50; see also Edwards 1986, 68, map, fig. 3, and App. A. 75. Both
publications omit the important evidence of Pecorella, Salvini et al. 1984, and the latter omits
references to several surveys to the south listed above.
21 Dyson 1973, 699; Sagona 1984, 122, 242243; Meyer 2001, 313, n. 43; more information
on Hasanlu is forthcoming in a publication by Dyson and Mary Voigt. For summaries of

excavations at Godin Tepe and surveys south of Urmia see: Swiny 1975, 7798; Young 1966,
228239; Young 1975, 191193; Howell 1979, 156157; Sagona 1984, 243ff., nos. 159178, and maps
V, VI, VII; also Meyer 2001, 311315 (on 315, n. 61, he lists 20 ETC sites from Howells survey, but
she gives only 10 examples).
22 Meyer 2001, 311, 313.
23 Trifonov 2000, 246, 261262. On the internet (August) Rsum of the Troisime Congrs
International sur l Archologie du Proche-Orient Ancien, Paris, April 1519, 2002 (given
to me by E. Izbitser), he gives the full range of Maikop as ca. 3700/35003200/2900 bc; in
the same internet site S. Korenevskii gives ca. 3950/36002900 bc, both based on corrected
C 14 analyses; Chernkh 1992, 7879, dates the Maikop kurgans to the 2nd half of the IVth
millennium bc, continuing into the IIIrd.
94 chapter three
and Maikop cultures coexisted in their respective homelands. But whether

such a situation existed in northwestern Iran is another matter, although it
remains an issue to be investigated. In any event, S Girdan is at present the
sole locus for investigation. And what objectively exists for investigation is:
the S Girdan tumuli are not ETC; the axes from these tumuli are not ETC;
neither the silver vessel nor scepter from Tumulus III is ETC. All (except the
scepter) have relevant parallels in the North Caucasus, Maikop culture. The
axes from Tianeti and Yerevan in the southern Caucasus (Georgia and Arme-
nia) do not per se contradict a Maikop period attribution; they could viably
document another example of a Maikop cultural southern expansionas
S Girdan.24 I therefore see it difficult to disagree with Trifonovs sequence
interpretation, otherwise one would have to posit two different cultures,
polities, and probably ethnic populations, co-existing for some time in basi-
cally the same geographical region. Only future excavations and analyses
will begin to clarify the issues.
Whatever chronological/cultural scenario eventually gets played out, in
any scenario the erection of the tumuli will have occurred sometime around
3000bc, perhaps more plus than minus. Eleven Maikop form tumuli, con-
taining Maikop period artifacts, were erected over several generations at one
locus. Their number reflects both a sustained use over time, and the con-
clusion that they served as the elite/royal necropolis for a substantial pop-
ulation, a self-conscious social and political entity. The two other Iranian
Maikop form tumuli mentioned above suggest that the cultures presence
may have extended geographically beyond S Girdan, in areas not neces-
sarily, on present information, later settled by a ETC population. Thanks to
Russian scholarship our recognition of the accurate cultural background of
the northwestern Iranian tumuli has revealed a major event in ancient Ira-
nian history, one that opens an important and complex period for renewed
research.25
24 R.M. Munchaev in his book (Munchaev 1975, 394) refers to his fig. 81, nos. 16, 17as axes
from Georgia. These are surely related to those from S Girdan but no information regarding
their proveniences or acquisitions are given: no. 16 seems closer (its long butt) to the Erevan
hoard examples than to those from S Girdan; see footnote 8 above.
25 I want to thank Karen Rubinson, Elena Izbitser, Alex Bauer, and Tony Sagona for biblio-
graphical information, and the latter two especially for valuable discussions on chronological
and cultural issues and implications.
Bibliography
Batchaev, V.M. 1984: Pogrebalne pamyatniki u seleniya Lenchinkai i Bylym. In:

Arkheologicheskie issledovaniya na novostroikakh Kabardino-Balkarii v 19721979
gg. 1 (Nalchik), 112163.
Batchaev, V.M., Korenevskii, S.N. 1980: Nakhodka originalnogo topora v maikop-
skom pogrebenii u s. Lechinkai. Kratkie soobshcheniia Instituta arkheologii AN
SSSR 161, 7983.
Burney, C., Lang, D.M. 1971: The Peoples of the Hills (London).
Chernykh, E.N. 1992: Ancient Metallurgy in the USSR (Cambridge).
Chechenov, I.M. 1984: Vtore kurganne grupp u selenii Kishpek i Chegem II. In:
Arkheologicheskie issledovaniya na novostroikakli Kabardino-Balkarii v 19721979
gg. 1 (Nalchik), 164253.
Deshayes, J. 1960: Les outils de bronze de lIndus au Danube (Paris).
. 1973: La Date des tumuli de S Girdan. Iran 11, 176178.
Dyson, R.H. 19: The Archaeological Evidence of the Second Millennium bc on the
Persian Plateau. In: Cambridge Ancient History. Vol. 11, 1, 3rd ed., 686715.
Edwards, M. 1986: Urmia Ware and its Distribution in North-Western Iran in the
Second Millennium BC Iran 24, 5777.
Howell, R. 1979: Survey of the Malayer Plain. Iran 17, 156157.
Kleiss, W. 1969: Bericht ber zwei Erkundungsfahrten in Nordwest-Iran. Archolo-
gische Mitteilungen aus Iran 2, 7119.
Kushnareva, K.Kh. 1997: The Southern Caucasus in Prehistory (Philadelphia).
Lyonnet, B. 2000: La Msopotamie et le Caucase du Nord au IVe et au dbut du IIIe
millnaires av. n.. In: C. Marro, H. Hauptmann (eds.), Chroniques des pays du
Caucase et de lEuphrate aux IVeIIIe millnaires (Istanbul), 299320.
Meyer, J.-W. 2001: Transkaukasus und Nordwest-Iran whrend der frhen Bronze-
zeit. In: J.-W. Meyer et al. (eds.) Beitrge zur Vorderasiatischen Archologie Wini-
fried Orthman gewidmet (Frankfurt am Main), 310321.
Munchaev, R.M. 1975: Caucasus In the Beginning of the Bronze Age (Moscow).
. 1994: Maikopskaya kultura. In: K.Kh. Kushnareva, V.I. Markovin (eds.), Epok-
ha Bronz Kavkaza i Srednei Azii: Rannyaya i srednyaya bronza Kavkaza (Mos-
cow), 158225.
Muscarella, O.W. 1969: The Tumuli at S Girdan A Preliminary Report. Metropolitan
Museum Journal 2, 525.
. 1971: The Tumuli at S Girdan: Second Report. Metropolitan Museum Journal
4, 528.
. 1973: The Date of the Tumuli at S Girdan. Iran 11, 178180.
. 1988: Bronze and Iron: Ancient Near Eastern Artifacts in the Metropolitan
Museum of Art (New York).
Pecorella, P.E., Salvini, M. et al. 1984: Tra lo Zagros e lUrmia (Rome).
Rezepkin, A.D. 2000: Das frhbronzezeitliche Grberfeld von Klady und die Majkop-
Kultur in Nordwestkaukasien (Rahden/Westf.).
Sagona, A.G. 1984: The Caucasian Region in the Early Bronze Age. BAR International
Series 214 (Oxford).
Swiny, S. 1975: Survey in North-Western Iran, 1971. East and West, 7798.
Trifonov, B.A. 2000: Maikop Type Tumuli in Northwestern Iran: Towards a more
96 chapter three
Precise Dating of the Tumuli at S Girdan. In: The Scholars Destiny, to the Cente-
nary of Boris A. Latynin (St. Petersburg), 244264 (in Russiantranslated for me
by Elena Izbitser).
Wiegartz, W. 1965: Die Ausgrabung am Tumulus Tepe Majid. Archaeologischer An-
zeiger 4, 788799.
Young, T.C. 1966: Survey in Western Iran, 1961. Journal of Near Eastern Studies 25,
228239.
. 1975: Kangavar Valley Survey. Iran 13, 191193.
chapter four
QALATGAH: AN URARTIAN SITE IN NORTHWESTERN IRAN*
Modern archaeological interest in Urartu and its culture has several phases.
The first, which lasted until around 1945, started in the 1870s, when, after
several objects reported to have come from Urartu appeared on the mar-
ket, the British Museum began to dig at Toprakkale in modern Turkey.
After a short time, D. Raynolds, an American missionary, and E. Clayton,
a British vice-consul, resumed work at Toprakkale for the Museum. Mean-
while, objects were continuously being purchased by various people from
local inhabitants, and these, together with the few objects and architectural
elements being excavated, formed a corpus of Urartian art. When in 1898
C.C. Lehmann-Haupt and V. Belck began what may be considered the first
attempt at scientific excavation of Toprakkalethe others could only be
called treasure huntsthey had a good idea what they expected. In 1911
1912, two Russian scholars, M.J. Marr and I.A. Orbelli, did some more digging
at Toprakkale, and at nearby Van, but interest soon petered out. Serious
excavation did not begin again until 19371938, when an American expe-
dition led by Kirsopp Lake undertook a campaign at both Toprakkale and
Van.
Russian scholars became interested in that part of ancient Urartu pres-
ently situated within the Soviet Unions borders, after 1930. At that time
survey teams began to seek out and record Urartian sites and inscriptions.
In 1939, B.B. Piotrovskii began one of the most significant excavations ever
undertaken in Urartian archaeology: the site of Karmir Blur (Red Hill), near
Erivan in Soviet Armenia. Work there continues to the present and one may
argue that the results achieved by Piotrovskii have played a large role in
reviving interest in Urartu.
We see, then, that British, American, German, and Russian archaeologists
were initially involved in the discovery of Urartus past.
The period after the Second World War begins the second, and more
intense, phase of Urartian scholarship. This phase is marked by very active
* This article originally appeared as Qalatgah: An Urartian Site in Northwestern Iran,
Expedition 13, nos. 34 (1971): 4449.

98 chapter four
archaeological campaigns led by Turkish scholars working within their bor-

ders. Intensive and extremely important surveys were also conducted most-
ly by British scholars, who discovered many Urartian sites and who made it
possible to define the borders of Urartu within Turkey. At the same time,
Russian archaeologists surveyed and excavated other Urartian sites and
cemeteries within their borders. American scholars played little or no active
role in Urartian archaeology at this time.
A third phase of Urartian archaeology began with recent discoveries
in northwestern Iran that made it clear, not only that Urartians set up
inscriptions in that area, but also, and more important, that they built cities
there.
In 1859 the Hermitage Museum received a group of Urartian objects said
to have come from Alishar (modern Shotlu) on the Iranian shore of the
Araxes River, near Mt. Ararat. And in 1905, at Guschi on the northwestern
shore of Lake Rezaiyeh, peasants found a hoard of Urartian objects in a
tomb. These objects, and a handful of scattered inscriptions, were all that
scholars could study if they were interested in the problem of Urartian
penetration into Iran. But it is of interest to note that no archaeologists seem
to have taken an interest in seeking out possible Urartian sites in Iran. In
1964 the situation began to change. The Hasanlu Project of the University
Museum and the Metropolitan Museum of Art, under the general direction
of R.H. Dyson, Jr., and the field direction of T. Cuyler Young, Jr. and the
author, excavated a small fort at a site we called Agrab Tepe, situated near
Hasanlu in the Solduz valley. Here we unearthed pottery that has good
parallels at sites in Urartu. While we are still not sure what the nature of
the fort was, we were aware at the time of excavation that the site had some
cultural contacts with Urartu. In 1968 the German Archaeological Institute
conducted the most important survey of northwestern Iran to date, and
Wolfram Kleiss, the leader of the survey, published his results in 1970. Kleiss
discovered about a dozen Urartian sites north of Shapur, and he began to
excavate one called Bastam, north of Khoy. In the same year Charles Burney,
who had worked in Turkey and who excavated an Urartian site there, found
evidence of Urartian culture at Haftavan, near Shapur. And again in 1968,
the Hasanlu Project team discovered the Urartian site of Qalatgah, which is
the subject of this article.
We look forward to continuous cooperation with our Iranian hosts and
colleagues in northwestern Iran. Joint work on the part of all scholars inter-
ested in ancient Iran will, we trust, generously expand our very limited
knowledge about both Urartian history in Iran and the nature of the mate-
rial and spiritual culture developed there.
qalatgah: an urartian site in northwestern iran 99
Red-slipped pottery vase from Agrab Tepe decorated

by burnishing and excised triangles filled with white
paste. Height, 56cm. University Museum Collection.
When travelling or excavating in the Near East, archaeologists often hear

from local villagers or workmen reports about the existence of stones with
writing, or the whereabouts of an ancient city. Too often the stone with
writing turns out to be a rough boulder covered with grooves and scratches
caused by weathering, or a weathered gravestone, or a millstone no longer in
use; and the alleged ancient city is actually a hill with outcropping rocks, or a
deserted Moslem cemetery. Nevertheless, one is sometimes given valuable
information about an object or site, and it is the duty of the archaeologist
not to be cynical after many false alarms, but to investigate every claim
about ancient remains made by local peoples. Thus, a few years ago T. Cuyler
Young, Jr. heard from his workmen at Godin Tepe in central western Iran
about a stone statue. As a routine matter he went to the place where the
statue was said to be and there found a stone stela of Sargon II of Assyrua
(722704 bc), thereby making an important historical discovery.
In July 1968, I was directing the Hasanlu Projects excavation of some
tumuli at S Girdan, situated near the modern village of Cheshm Gl in
100 chapter four
View looking north toward Qalatgah (rising to the

right behind the trees) as seen from S Girdan.
the Ushnu valley of northwest Iran. For several days after the beginning
of excavations, workmen pointed to a cluster of trees to the northeast, on
the slopes of a mountain that formed the eastern boundary of the Ushnu
valley. There, they said, was a very important place, an ancient city, and that
the area was called Qalatgah, or place of the fortress. On July 23 one of the
local landlords also talked of the site, and he added that there was a road
there and that in 1967 local peasants had found a large stone, broken into
three pieces, all of which had writing engraved. His description of stick
like writing suggested that he had seen cuneiform. Yes, he had himself
seen the stones and he was able to tell us that one or two pieces had been
taken by a local antiquities dealer (who never paid the promised money),
who in tum sold the stones to a dealer from Tehran; the third piece was
taken to the authorities in Ushnu, the valleys chief town. He knew the
Western edge of the Qalatgah site lies on lower slope

of nearby mountain (left). The seal was found along with
sherds in the low rise in the right center of the photograph.
exact place where the stones were found and would be glad to take us there.
At last on July 30, I first visited Qalatgah, along with Agha Z. Rahmatian, the
representative of the Archaeological Service of Iran, and Carol and Christo-
pher Hamlin, at that time graduate students at the University of Pennsylva-
nia.
Qalatgah is a large and steep site consisting of several high spots. A few
hundred meters up from the modern road, two magnificent springs gush
forth from a vertical rock outcrop situated behind a cluster of willow trees.
The remains of one or more buildings are still visible on the surface to the
north of the springs but we do not know their date. Many thousands of
stones litter the site over a wide area, and in at least two places on the slopes
to the northeast of the springs are stretches of fortification walls formed of
large, well-cut boulders, running east-west. (These walls were the road of
the landlord.) Further up the steep slope holding the walls, we found a level
area at the top and evidence of more walls just protruding from the surface.
From this highpoint we could see the whole Ushnu valley and a great part
of the Solduz valley to the east, as far as the modern town (and ancient Iron
Age tepe) of Nagadeh.
102 chapter four
West slope of the main site of Qalatgah showing

remaining stone blocks (in situ) of the fortification wall.
On the eastern slope of the site, not visible from the Ushnu side, there
is a large open rectangular rock chamber, apparently natural in origin but
showing signs of having been worked by man; the open area is now used as
a shelter by shepherds. In his recent article Urartische Pltze in Iranisch-
Azerbaidjan Wolfram Kleiss reports that at the Urartian site of Kale Waziri,
at the northwest corner of Lake Rezaiyeh, there is a Felshle, an open rock
chamber, that was natural in origin but showed signs of human working.
Surely these two chambers had similar functions, which may be known to
us after Qalatgahs excavation and more research.
Of the many sherds we picked up over an extensive area of Qalatgah,
all appear to be Iron III types (post-9th century bc), painted wares with
plain or hatched triangles decorating the inner rim of cups and bowls and
monochrome plain buff and red-burnished wares, well known from nearby
Hasanlu in Period III. Hollow-based bowls, simple, outward curving plain
bowls, carinated rim bowls, and cups, are among the shapes represented by
Painted triangle ware sherds from (a) Qalatgah, (b) Ziwiye, (c) Hasanlu.
the Qalatgah sherds. Similar inner-rim painted triangle decoration has been
found at such sites as Hasanlu and at Ziwiye farther to the southeast, and in
Urartu at Altintepe, in the final upper level, and in the Van area at Van itself
and at nearby Tilke Tepe. The red-burnished wares may also be related to
similar types of ware found in several Urartian sites. Moreover, in addition
to the painted-triangle ware and red-burnished ware sherds, at home in
northwest Iran and Urartu, we also found some very fine, highly polished
redware sherds known to archaeologists as Toprakkale Ware, so-named
after the site in Urartu where they were first found, but since recognized at
other Urartian sites, and also recently found by C. Burney at Haftavan Tepe,
near Shapur in Iran.
A second trip on August 9th to the site with the whole staff of the Hasanlu
Project proved to be more rewarding. Not only did we find more diagnostic
sherds of the types just discussed, but I found a white stone stamp-cylinder
seal of characteristic Urartian type and motif. More important than this was
the find made by Christopher Hamlin: an Urartian inscription on a broken
104 chapter four
Urartian seal of white stone found on the surface of Qalatgah.
Urartian inscription of kings Ishpuini and

Menua, from the surface at Qalatgah.
stone block. The stone was found, together with other plain stones, built
into a modern dam holding back a pool of water, fed by the two springs of
Qalatgah, and used for local irrigation.
The seal is a concave cylinder, free of decoration, with a stylized horned
animal running to the right carved at the base; a loop for suspension is at
the top. Seals of this type have been called stamp-cylinders by R.D. Barnett,
and they have been excavated at Karmir Blur and Toprakkale in Urartu, and
at Igydr and Kelankran in the southern Caucasus, made there, no doubt,
under Urartian influence. I believe the example from Qalatgah is the first
from northwest Iran.
Although they have been handicapped both by the weathering of the
stone and by its fragmentary state, Christopher Hamlin and Maurits van
Loon have been studying the inscription. Van Loon has sent me a prelim-
inary translation which I am quoting here; he will shortly publish a more
extensive commentary and discussion in the Journal of Near Eastern Stud-
ies:
1) When, relying upon the god Haldi and upon the god T[eisheba?, Ish-
puini,]
2) son of Sarduri, king of Urartu [and Meinua.]
3) son of Ishpuini, of the country Sapaya []
4) the king? they too [k? ] both the god Hal[di and ]
5) for the god Haldi they change[d to hi]s? city Uishe? of the country []
6) to the trees of they carried
The text, therefore, had been set up for public viewing by Ishpuini and
his son Menua, sometime during the co-regency of these kings, that is
sometime around 810 to 805bc. It should also be recalled here that during
the co-regency of Ishpuini and Menua they set up an inscription written in
both Assyrian and Urartian within the Kel-i Shin pass. That inscription is
therefore from the same period as this Qalatgah inscription found in 1968
and published above.
In the 1969 issue of the Archaologische Mitteilungen aus Iran, J. Friedrich
published an incomplete stone inscription found in the Ushnu valley in
1967. Friedrich says the fragment, now in the Tehran Museum, is one of
two stones recovered and that the other is in the hands of an antiquities
dealer. I suggest that there can be little doubt that this fragment is one of the
inscribed stones reported to me as being found at Qalatgah in 1967, the find-
spot of which we were seeking when we began exploring Qalatgah. Whether
Friedrich, or his source, or the Cheshm Gl landlord who mentioned three
stones, is correct, is not presently known. Friedrichs inscription was written
106 chapter four
for Menua, son of Ishpuini, and records the erection of a special building as
well as a city, not named in the fragment. Therefore, this inscription was set
up by Menua after the termination of the co-regency period with his father
Ishpuini and is thus to be dated between about 805 and 786 bc, contempo-
rary with a well-known inscription set up by Menua at Tash Tepe some fifty
miles to the east of Qalatgah. The Tash Tepe inscription mentions the found-
ing of a city in the area (not yet discovered), and is farther southeast than
any other inscription of Urartian origin presently known.
The Qalatgah inscriptionsthe one found by the Hasanlu Project team,
and the one published by Friedrichdocument the historical fact that
by the late ninth century bc and the earliest years of the eighth century
an Urartian city existed in the Ushnu valley, close to the Kel-i Shin pass.
The Tash Tepe inscription further establishes that Menuas penetration into
northwest Iran extended east across the Solduz valley and the southern
shore of Lake Rezaiyeh. In short, this inscription, taken together with Kleiss
recent discoveries, enables us to state that Iranian territory west and south of
the lake was under Urartian control at the beginning of the eighth century bc
The pottery found at Qalatgah cannot yet give us exact information about
the duration of the site because there is still much work to be done in sorting
out the sequences of Iron III pottery both in northwest Iran and in Urartu.
Kutlu Emre, a Turkish archaeologist, has recently noted that monochrome
burnished wares existed throughout the history of Urartian cities, that is
in the eighth and seventh centuries bc. The painted-triangle ware sherds
from Qalatgah are the same as those found at Hasanlu IIIB and at Ziwiye.
Also, they appear to be very close to, or the same as, the painted pottery
from Altintepe in western Urartu, there found only in the upper level, not
in the lower where only monochrome pottery was in use. The lower level
is dated by the excavator T. zg to the eighth century bc (the founder of
the city may have been Arghisti II, about 714685 bc), the upper level to the
second half of the seventh century bc, a date consistent with the terminal
dates suggested for Hasanlu IIIB and Ziwiye (new C-14 adjusted dates for
the terminal date of Hasanlu IIIB average to 679bc with a half-life of 5730
years). Painted-triangle ware is also reported at Van and nearby Tilke Tepe,
as already stated, but the exact relationship of this ware to the monochrome
ware found at these sites is not clear to me, as it is stated by some archae-
ologists (including the excavator K. Lake) that the painted ware was in use
earlier, during the eighth century, than the monochrome ware, dated there
to the seventh century. If true, the situation would be exactly opposite the
pottery sequence reported at Altintepe. Moreover, at Toprakkale, which is
close to Van and Tilke Tepe, monochrome but not painted pottery is found.
Toprakkale was apparently founded by Rusa II (685645bc), which if true

would mean that painted pottery was not in use there during the seventh
century. At Karmir Blur in the northeast, a city also apparently founded by
Rusa II, monochrome pottery is all but universal, painted pottery being rare.
Again there is a difference to the pottery sequence reported at Altintepe.
Kutlu Emre suggested that the present excavations at Van and Patnos may
help us understand the relationship of painted and monochrome pottery,
and we look forward to published reports on these sites for further clarifica-
tion of the problem.
In Anatolia, Phrygia, painted pottery was in use during the late eighth
century bc and continued into the seventh after the Cimmerian invasion.
In central western Iran at Sialk and in Luristan, painted pottery was in use
probably by the late ninth century, almost certainly during the eighth, if, of
course, one accepts the pottery of the Sialk B cemetery to be late ninth and
eighth century bc in date, or at least eighth century, as I do believe; but the
origin of painted pottery in central western Iran is a complex question and
need not concern us here.
The site of Qalatgah has not yet been excavated. Nevertheless a few hours
of careful surveying, carried out as part of the Hasanlu Projects goal, to
uncover as much information as possible about the history of the Ushnu
and Solduz valleys, have turned up some important historical data, hitherto
unknown. All we are able to say, and that tentatively, is that it would seem
that Qalatgah continued as a site until it was finally destroyed or abandoned
sometime during the seventh century, having had an existence of a hundred
or more years. The seventh century witnessed major upheavals and move-
ments of peoples in the area, and there were many wars between Assyrians,
Scythians, Urartians, and Iranian tribes. Whether Qalatgah remained under
Urartian control throughout its history or changed hands is not known to us
at present; excavation may make it possible for us to answer the question.
Studies of ancient geography are ventures into the unknown and much
difference of opinion exists. All one need do is compare the various maps
published by different scholars, each of whom assigns different ancient sites
to the same geographical regions. Therefore, it would be premature and
dangerous to speculate at present that the ancient city of Qalatgah was
built in the land of Barsua, or Manna, or in a Mannean-controlled area
such as Uishdish or Subi, states known to us from Assyrian records. At
the same time, it would certainly now be incumbent on those concerned
with ancient geography to take into account the archaeological evidence
of Urartian control over the area west and south of Lake Rezaiyeh in the
eighth century bc. With this thought in mind, I throw out for discussion a
108 chapter four
tentative suggestion with respect to one aspect of ancient geography that

could emerge from our new knowledge about Urartian control in Iran.
The great French scholar Thureau-Dangin published the text of a letter
describing Sargon IIs eighth campaign that took the king into Iran, and,
according to Thureau-Dangin, into Urartu, before he returned south to sack
the Urartian temple site of Musasir, somewhere in northern Iraq. The text
mentions the many rivers, mountains, cities, and regions passed by Sargon
on his campaign. Thureau-Dangin had worked out the route and made a
map to illustrate it. His map shows Sargon travelling north around the east
shore of Lake Rezaiyeh, then turning west into Urartu where he went around
Lake Van before returning south, to Musasir and later home. Sargons text
does not, however, mention going around any lake, and it has always seemed
strange to me that Sargon did not mention Lake Rezaiyeh (a large lake) if in
fact he passed by it, especially when his text is quite explicit with regard
to geographical features. Sargon also only casually refers to Urartian cities
on the shore of a sea, without naming that sea. Thureau-Dangin and most
scholars have taken that sea to be Lake Van because, one assumes, of the
mention of Urartian cities. But must it be Lake Van? Is it not possible that
Sargon never left northwest Iran and that he referred to those Urartian cities
we now know existed near Lake Rezaiyeh when he talked of attacking Urar-
tian cities by the sea? If possible, Sargon would then have gone to Musasir
from the Lake Rezaiyeh region via one of the several passes existing in the
Zagros connecting Iran and Iraq. Indeed there are problems connected with
this suggestion, one of which concerns the use by the Assyrian scribes in
several texts of the term Upper and Lower seas: which term refers to
Lake Rezaiyeh? However, the suggestion presented here might be tested by
those scholars more knowledgeable than I in matters concerning ancient
geography; and if the view presented here about Sargons campaign meets
with approval, then the ancient geography of northwest Iran will have been
altered drastically and we will have to reconsider our opinions about the
geographic locations of Parsua, Zamua, and other neighboring states. In
any event, an analysis of Sargons route during his eighth campaign cannot
any longer ignore both the absence of any mention of the circling of Lake
Rezaiyeh, and the presence of Urartian cities along its southern area, and
also west of the lake, as reported by Kleiss.
chapter five
EXCAVATIONS AT AGRAB TEPE, IRAN*
Agrab Tepe is one of three mounds lying close to the modern village of
Dalma in northwestern Iran (Figure 1), southwest of Hasanlu in the Solduz
Valley near the low ridge that separates the Hasanlu plain from the mod-
ern town of Nagadeh. Of the three mounds, one called Dalma Tepe was
excavated in 1961 and yielded Neolithic remains.1 The third mound is still
unexcavated.
The distance between Hasanlu and Agrab is approximately two miles,
about a forty-five-minute walk from one site to the other; the sites are also
visible to each other, but the low ridge to the south cuts off the view of the
Nagadeh plain from Agrab. A modern road that connects Nagadeh to points
north passes by Agrab about a half-mile to its west; it cannot be established if
this road follows an ancient track. Before excavation the mound was about
52 meters in diameter and about 6 meters in height. A spring that caused
the surrounding land to be swampy and impassible in 1964 exists about 100
meters to the southwest. The site was built directly over a rock outcrop, the
only one visible in the area (Figure 2).
In 1964 the Hasanlu Project, a joint project of the University Museum
of the University of Pennsylvania and the Metropolitan Museum of Art,
decided to excavate a second mound at Dalma as part of its ongoing plan to
collect archaeological and cultural data of the Solduz Valley, south of Lake
Urmia (Rezaiyeh). Work at Hasanlu had to be suspended while we worked
at Agrab, and we were able to devote a total time of three full weeks to the
completion of our task. The mound had no local name. Since it was thought

1 T. Cuyler Young, Jr., Taking the History of the Hasanlu area back another five thousand
years , ILN (Nov. 3, 1962) pp. 707 ff.; R.H. Dyson, Jr. Excavating the Mannaean Citadel of
Hasanlu , ILN (Sept. 30, 1961) pp. 534 f. References to Agrab Tepe: Dyson 1965, p. 212f., pl.
xliii, fig. 12; R.H. Dyson, Jr., Hasanlu and the Solduz Valley, Archeologia Viva 1 (1968) p. 85;
Young 1967, p. 30; W. Kleiss, Zur Ausbreitung Urartus nach Osten, Istanbul. Mitt. 1920 (1967
1970) pp. 127, 129; also his Bericht ber Zwei Erkundungsfahrten in Nordwest-Iran, AMI 2
(1969) p. 26.
110 chapter five
Figure 1. View north from south ridge, past Dalma village. At center,
Agrab Tepe. To the right, in the plain, Dalma Tepe. The village
of Shetanabad, just west of Hasanlu, is at the upper right.
Figure 2. Agrab Tepe from the south, after completion of the excavations.
that the use of the name Dalma No. 2 would cause confusion with respect
to the Neolithic mound, it was decided to refer to the site as Agrab Tepe.2
Agrab Tepe consists of a single large building built over a rock outcrop
(Figures 3, 4). The building is a fortified structure with thick outer defensive
2 Scores of scorpions were killed, hence the name Scorpion Mound. The staff consisted
of T. Cuyler Young, Jr., and the author as co-directors, Louis D. Levine and Ted Rathbun as
archaeologists, and Ed Keall, architect. The director of the Hasanlu Project, whose advice in
the production of this report I here acknowledge with thanks, was R.H. Dyson, Jr. I also take
pleasure in expressing my thanks to T. Cuyler Young, Jr., and Louis D. Levine for commenting
critically on many items discussed in this report. For a good map of northwestern Iran, Kleiss,
Bericht tiber Zwei Erkundungsfahrten, fig. 1.
excavations at agrab tepe, iran 111
Figure 3. Plan of Agrab Tepe.
walls whose interior surface also functioned as walls of the rooms. The
defensive wall consists of a foundation of large, roughly cut stone blocks
with small stones used for chinking (Figures 5, 6, 9), and with a brick super-
structure. Eight buttresses or piers and a massive tower-entrance flanked
by two piers project from the walls. In plan the structure is irregular, look-
ing like a flattened oval at one end, with no obvious compass orienta-
tion. It measures about 31 meters east-west and about 28 meters north-
south.
The walls are 1.90 to 2.00 meters thick, except for the two units east of
the entry tower; the first unit varies from about 1.80 to 2.00 meters, the
second from about 1.50 to 1.65 meters. The eight piers are irregularly spaced,
varying from 4.5 to 5.7 meters distance from each other. Their width also
varies from 3.00 to 3.70 meters; and their projection from the wall varies
from 1.10 to 1.30 meters. If the structure were not so well made one could
assume some haste in the building activities reflected by these irregular
measurements.
112
chapter five
Figure 4. Sections of Agrab Tepe.

Figure 5. Tower-entrance, shaft, and blocked door, facing north.

114 chapter five
Figure 6. Tower-entrance shaft, with bedrock, facing north.
The entrance unit is a tower 9 meters in width, projecting 7 meters from

the defensive wall. Its south wall is 4.30 meters thick, and is preserved to a
height of five courses of stone measuring 3.70 meters on its outer side and
3.00 meters on its inner side, where it rests on the rock surface (Figures 4,
6). Originally the tower must have been at least seven or eight courses and
another meter in height, to judge by the preserved height of the defensive
wall. The side walls are 2.00 meters thick, becoming 4.20 meters thick when
joined to the corner piers; the latter begin 5.10 meters in from the front of
the tower.
Within the tower there is a shaft measuring 5.5 3.30 meters, and 3.40
meters deep to bedrock on the north end (Figures 3, 4, 5, 6). A mud-brick
wall .70 wide and 1.90 meters in height divided the shaft into two areas.
No obvious means of entry into the shaft exists, and presumably a ladder
was employed. Nor is there any evidence available to suggest how one
entered the tower from the plain below. Presumably a door existed in the
now destroyed brick superstructure, and one has to assume that a ramp, or
ladder, probably portable, allowed access from the plain below. No other
primary function for the tower other than that of an entrance unit, or
gateway, comes readily to mind.
Directly in the center of the defensive wall facing the tower, the north
wall of the shaft, was a doorway 1.25 meters wide, with a large stone used as
a sill (Figures 5, 6). The door was eventually filled in with stones and a new
sill was built at what seemed to be level 2 (counting from the bottom up)
of room A2/B6. Whether this door served only as a passage to and from the
tower, or served also in some manner as an entry to the shaft, is not clear.
In any event, there is no other indication of a passage from the tower to the
main structure.
The structure as preserved consists of thirteen rooms or areas, most of the
walls of which are constructed of brick, sometimes set on a single course, 10
to 20cm. in height, on stone foundation, other times set right on the floor
surface. In rebuilding, sometimes a stone foundation layer was placed over
an earlier wall stub, other times the new wall was placed directly on the stub.
Brick size is uniform throughout the structure, 10/12 49/50 49/50 em. The
walls were coated with a thick layer of mud plaster.
A rectangular room, A2/B6, 5.30 3.60 meters, led to most of the other
rooms and to the stairway area to the west. To the right is a small room,
B10, 3.30 3.00 meters. In Period 2 it had a drain constructed of a small
sunken pithos surrounded by flat stones; this was connected to a draining
system originating in B4 (Figure 7). During Period 2 the doorway of B10 was
narrowed by the addition of two stubs of brick.
Abutting B10 is a small triangular room, B8, 3.10 2.75 meters. In Period
1 this room had a door in its northeast corner (not shown on the plan)
that entered into the area called B5. During the rebuilding of Period 2 this
doorway was blocked. B8 was either abandoned in Period 2 or was entered
from above.
West of room A2/B6 is a rubble-surfaced area, A3, enclosing a rectangular
brick pier that must be interpreted as a support for a stairway. A3 flanks the
stairway support on three sides and continues up to the fortification wall
at the west; the area south of the stairway support was unsurfaced and had
a round terracotta hearth in Period 1. Except for those rooms entered from
above by a ladder, one presumably had to use this stairway to communicate
between stories, and perhaps to reach the battlements. The stairway was
116 chapter five
Figure 7. Draining system in B10. The wall has been cut away.
entered directly from A2/B6, which in turn served as a passage to the other
rooms. One is here reminded of the entrance system used at Hasanlu IV,
where stairways were placed in a room to one side of the anteroom.
Directly to the north of room A2/B6 is room B3, 4.20 3.60 meters. It is
connected by doors to rooms D1 and D2. A curious stone-edged semicircular
stepdown, about .25 deep and 1.25 meters in diameter, exists in the rooms
southeast corner, taking up about half the rooms space (Figure 8). Nothing
was found to suggest what its function might have been.
Room D1 is triangular in shape and abuts onto the defensive wall (Fig-
ures 8, 9); its two walls are each about 3.50 meters in length. This area no
doubt functioned as a kitchen and storage room, since many animal bones
and occupational trash were found in the fill and on the floor; also, five pithoi
Figure 8. Room D3, D1 beyond.
were excavated in situ resting on the floor (Figures 3, 8, 9), four of these
against the defensive wall, and sherds of others were found in the fill.
Room D2, 2.50 1.85 meters, also abutted onto the defensive wall (Fig-
ures 10, 11). It apparently functioned as a storeroom, for a large pile of stones,
presumably slingstones, was found resting in the northwest area against the
defensive wall. Remains of a late wall, apparently Period 2 or 3, ran north-
south; the slingstones were found partly under it, resting on the primary
floor. The south wall of D2, seen on the plan as a double wall, actually con-
sists of several wall periods.
Room C1, 4 8.5 meters (Figures 10, 11), abutted against the defensive wall.
In Period 1 a long northsouth wall (no door is visible in the low wall stub,
but it could have eroded away) divided the area from D2 and continued
118 chapter five
Figure 9. Room D1, facing east.
further south. In Period 2 a wall running along the line of the earlier wall, but
slightly displaced to the east, was built; only a stump, which projects into C1,
now remains (the stone foundation is visible in Figures 10, 11); presumably
it continued to the defensive wall. In the fill and on the floor of C1 were
found charred grain, stone pounders, several smashed pithoi, and sherds,
suggesting a work and storage room.
The rooms to the south of C1 are the most unusual of the structure. A large
rectangular room, 6.00 9.00 meters, was subdivided into four rectangular
rooms of unequal size by a cross-wall of brick resting on a stone foundation
(Figures 3, 4, 12). The cross-wall was built against a well-made outer wall,
70cm. thick, constructed of small stones, which in turn was built against the
brick walls of the neighboring rooms, and also against the defensive wall.
Figure 10. Room C1, D2 beyond, facing west.
This stone wall was constructed as one unit before the subdivision of the
area. In addition, the floors of the whole area were paved with flat stones
one layer thick. A layer of earth about 15 to 20cm. thick separates the paved
floor from the foundations of the cross-wall (Figure 12); it is therefore clear
that the paved room with its stone-lined walls existed as one large area
for a time before it was subdivided. The floors of the four rooms now all
slope slightly toward the center of the area. It is not clear if this represents a
sagging (presuming the flooring was not laid directly on the bedrock) or an
original plan. A stone-lined and capped draining system, 3.30 meters long,
25cm. high, and 15cm. wide, was constructed through both the southwest
corner of the stone wall in B4 and the neighboring brick wall. It emptied
into the drain in the northwest corner of B10 (Figure 7); no drains existed
120 chapter five
Figure 11. Room D2, C1 beyond, facing east.
in the other stone-paved rooms. Presumably the drain was built before the
subdivision occurred.
The particular time in the history of the building when the area was
given a stone lining, was paved, and was subsequently subdivided seems
fairly clear. The stone lining wall was built against the brick wall of Period 2,
which itself rests on the stub of the earlier Period 1 wall, and which blocked
the doorway entering into B8. Therefore, the lining and paving belong to
the second construction period and the subdivision to a later phase of this
period. No doorways exist to connect the four rooms to each other or to the
neighboring rooms, nor is there any indication that the abutting brick walls
once had doors. Thus here, as with rooms C1 and B8, entry was doubtless
from above. This feature at Agrab plus the stairway makes it certain that
another story existed over the level preserved to us.
Figure 12. Rooms B1, 2, 4, 5, facing east.
The nature of this particular area within the Agrab structure remains
a mystery. Surely the elaborate walling, paving, partitioning, and draining
system reflect a function not shared by the other rooms. Was it originally
built as a bathing area? Could it have been built as a rodent-proof, moisture-
free storage room? Pithoi fragments were found in Room 1, and animal bones
were found in the fill of areas 1, 2, and 4, but these could have fallen from a
higher story. The problem remains unresolved.
Several test trenches were dug outside the defensive wall, in the north,
south, east, and west. Nothing but bedrock was encountered in three of
these trenches, but in the east trench a small enclosed space was excavated.
It was formed by two short brick walls projecting from the front of the two
piers, and creating an open doorway 1.75 meters wide. The walls are two
122 chapter five
Figure 13. Section of room D1.
bricks wide, or from 1.00 to 1.10 meters thick. The sill was also of brick and
exited to the bare rock below. There was no visible means of communication
between this space and the main unit, nor was there any visible means of
closing the door in the space. Three arrowheads, a bone spatula, a bead,
and a grinding ball were found here in the fill. Perhaps the unit served as
a temporary postern-gate area.
The building at Agrab Tepe was destroyed at least twice by fire and rebuilt
using the same basic plan, thus creating three periods. During most of
the course of excavation this fact was not recognized for several reasons.
The rebuilt walls were placed directly over the earlier walls (Figures 810).
The heat generated by the fires was quite intense and vitrified the plaster,
which made it difficult to clear wall faces. This situation prevented us from
seeing wall stratification and offsets of later walls over earlier ones, which
occurred in a few cases. Moreover, in only a room or two were there any
floors preserved from Period 2, as these floors were not hard, nor were
they regular (Figure 13). Thus we assumed that we were digging a site with
one occupational level. When we were able both to read the sections and
examine the walls closely, we concluded that Agrab had several levels. By
this time the digging was basically finished and it became difficult in some
cases to divide the pottery into the three levels. However, Period 3 was close
to the surface and had no recognizable floors or remains as such, only sherds
considered to be surface finds. An Islamic level, much denuded, had been
cut into it and destroyed it. Also, many of the finds were in the fill over the
primary floor level and presumably came from Period 1. Yet, one cannot be
sure in all cases. In room D1 several pithoi were found on the primary floor,
supplying us with important information.
The lack of good, firm floors in Period 2, plus the rebuilding on original
plans, suggests that only a short amount of time elapsed after the first
destruction before rebuilding, and that it too may have been rapidly rebuilt
and only briefly inhabited in Period 3.
Objects Excavated
Pottery
Bowls with rolled rims: Figure 14, 1 and 2, plus more sherds; coarse ware,
buff, lightly burnished. Related bowls come from Hasanlu IIIB and A (Young
1965, fig. 6, 5); Bastam (Kroll 1970, fig. 6); Godin II (Young 1969, fig. 44, 13,
15); Baba Jan A (Goff Meade 1968, fig. 10, 4); see also Zendan (Boehmer 1961,
pl. 60,5; pl. 55, 14).
Carinated bowls with square rims: Figure 14, 3, two examples; coarse
ware, buff, unburnished. Related bowls are from Hasanlu IIIA;3 Godin II
(Young 1969, fig. 43, 14); Pasargadae, unpublished; Geoy Tepe A (Burton-
Brown 1951, fig. 36, 357).
Pots with oblique shoulder spouts: Figure 14, 4, with handle, 5; three
more examples, one with handle; coarse ware, buff, unburnished. Related
pots are from Hasanlu IIIB (Young 1965, fig. 2, 5); Geoy Tepe A (Burton-
Brown 1951, fig. 35, 126; fig. 40, 1644; fig. 41, 113); Godin II (Young 1969, fig. 42,
17); Norsuntepe (Hauptmann 1970, fig. 16, 12; fig. 17, 3; fig. 18, 1, earliest Iron
Age level); cf. Altintepe, later level (Emre 1969, pl. v, 1); Achaemenid Village
(Ghirshman 1954, pl. XXIX, G.S. 1206b, G.S. 959).
Pots with one handle: Figure 14, 6 and 7; coarse ware, buff, unburnished.
Similar vessels come from Hasanlu IIIB; Geoy Tepe A (Burton-Brown 1951,
fig. 35, 106).
3 When Hasanlu IlIB and A appear without a reference it means that I found the vessel
in the Hasanlu files of the University Museum.

124 chapter five
Figure 14. Pottery from Agrab Tepe.

Lip spout: Figure 14, 8, one example; buff, smoothed surface, medium grit
interior. Related spouts are from Hasanlu IIIB and A (Dyson 1965, pp. 205, 212,
note 36); Geoy Tepe A (Burton-Brown 1951, pl. 39, 219); Zendan (Boehmer
1961, pl. 45, a1; pl. 51, 21); Ziwiye, unpublished (Young 1965, fig. 10 chart);
Achaemenid Village I (Ghirshman 1954, pl. XXIX, G.S. 2242); Luristan (Goff
Meade 1968, p. 123, fig. 11, 14, Iron III; Vanden Berghe 1967, pl. 59, no. 2);
Masjid-i-Suleiman (Ghirshman 1970, p. 184, pl. IVb).
Plain bowls with in curving sides, pinched rims: Figure 14, 9; one buff,
red-slipped, burnished with fine paste; two others, without hollow base,
buff, smoothed surface. Related bowl shapes are known from Hasanlu IIIB
and A (Young 1965, fig. 1, 1); Ziwiye (ibid., fig. 3, 1); Bastam (Kroll 1970, fig. 1, 1;
pl. 2, 1, 2); Godin II (Young 1969, fig. 43, 2); Achaemenid Village II (Ghrishman
1954, pl. XXXVII, G.S. 1219f); Norsuntepe (Hauptmann 1970, fig. 23, 1, middle
Iron Age); Armavir Blue (Barnett 1963, fig. 19, bottom). Hollow-based bowls,
not to be confused with omphalos bowls, occur at Agrab in the earliest
period and in the fill. They also occur at Hasanlu IIIB (Young 1965, fig. 1,
2); S Girdan (Muscarella 1971a, fig. 29); Qalatgah surface (Muscarella 1971b,
p. 46); Godin II (Young 1969, pl. 44, 6, 7; also earlier in Godin III, fig. 32, 7, 8);
Altintepe, earlier level (Emre 1969, p. 295, fig. 12); cf. also Karageorghis 1962,
p. 114, and pls, 144, 145, 148, 149, 156, 166, 171, dated to the early sixth century
bc. The hollow base seems to be a variant of the omphalos, which occurs in
Iron III also. Bowls from Igdyr (Barnett 1963, fig. 15) may be hollow based,
rather than omphaloi.
Bowls with wide flaring collar and pinched rim: Figure 14, 11; one
example is coarse, buff, unburnished, another is buff, with a smoothed
surface, a third is burnished buff orange ware, with a fine paste. Similar
bowls occur at Hasanlu IIIA; Ziwiye (Young 1965, fig. 3, 11; cf. 6, 9, 12, with
omphalos); Zendan II (Boehmer 1961, pl. 50, 7, 8, with omphalos; Boehmer
1967, pp. 577, 580, fig. 8, AC; also fig. 8 for Ziwiye and for 7th-century Nimrud
examples); Yanik Tepe (Burney 1962, pl. XLV, 36, but painted); Qalatgah
surface (Muscarella 1971b, p. 47); Altintepe, later level (Emre 1969, p. 299,
pl. v, 1); Norsuntepe (Hauptmann 1970, fig. 21, 4, middle Iron Age); Persepolis
(Schmidt 1957, pl. 72, 1); Pasargadae, unpublished; Gordion, Persian level
(R. Young 1962, pl. 41, fig. 1a, b); Samos, wood (Kopcke 1967, p. 119, fig. 6, 7th
6th century bc.)
Carinated bowls with pinched or slightly rolled rim: Figure 14, 10

and 12, hollow base, Figure 15, 1 and 2; these vessels occur in a variety of
126 chapter five

surfaces: coarse ware buff; burnished buff; coarse buff, slightly burnished;
buff, smoothed surface; two (like Figure 14, 12 and Figure 15, 1) are red-slipped
buff; one (like Figure 15, 2) is gray burnished. Similar bowls are found at
Hasanlu III B and A (Young 1965, fig. 1, 2, 4; fig. 2, 6; cf. also Hasanlu IV, fig. 6,
2, 4); Bastam (Kroll 1970, fig. 1, 2, 8); Ziwiye (Young 1965, fig. 3, 14); Zendan I, II
(Boehmer 1961, pl. 56, 2325); Godin II (Young 1969, fig. 44, 17, 18); Haftavan
(Burney 1970, p. 170, 6); Baba Jan I (Goff 1970, fig. 8, 7); Armavir Blur (Barnett
1963, fig. 19, top); Patnos (Ankara Museum); Kef Kalesi (gn 1967, fig. 16);
Van (von der Osten 1952, pl. VIII, 3, 4).
Bowls with everted rims: Figure 15, 3; red-slipped buff ware. Similar bowls
occur in Hasanlu IIIA; Ziwiye, unpublished; Zendan (Boehmer 1961, pl. 50,
9); Godin II (Young 1969, fig. 43, 3); Yanik Tepe (Burney, 1962, pl. XLV, 33,
painted); see also Igdyr (Barnett 1963, fig. 16, bottom) red polished; and Van
(von der Osten 1952, p. 325, pl. VI, I).
Bowls with inward-curving rolled rims: Figure 15, 4, coarse, buff ware;
5, red-slipped, burnished; 6, coarse, buff; 7, red-slipped, burnished; others,
like 4 and 5, red-slipped, burnished. Such bowls are known from Hasanlu
IIIB and A (Dyson 1965, p. 204, fig. 13; Young 1965, fig. 1, 3, 5; common in IIIB);
Ziwiye (ibid., fig. 3, 19); Bastam (Kroll 1970, fig. 1, 10; fig. 3, 1, 7; fig. 4, 3); Baba
Jan I (Goff Meade 1968, fig. 8, 9); Norsuntepe (Hauptmann 1970, fig. 23, 3, 4);
Altintepe, earlier period (Emre 1969, fig. 8, 10; pl. II, 1, 2); Van (von der Osten
1952, pl. VIII, 1).
Bowls with rolled and grooved rims: Figure 15, 8, burnished gray ware,
fine paste; 9, buff, smoothed surface. Similar shapes occur at Hasanlu IIIA;
Ziwiye (Young 1965, fig. 3, 4, 5); Godin II (Young 1969, fig. 43, 11).
Carinated bowls with grooved collars: Figure 15, 10, buff, smoothed
surface; 11, burnished gray ware. Similar forms occur at Hasanlu IIIB; Geoy
Tepe A (Burton-Brown 1951, fig. 36, 643); Godin II (Young 1969, pl. 43, 12);
cf. Norsuntepe (Hauptmann 1970, fig. 18, 4; fig. 21, 3); Achaemenid Village
(Ghirshman 1954, pl. XXXVII, G.S. 1219g).
Carinated bowls with slightly rolled rims: Figure 15, 12 (64-4), 13 (64-3)
from fill over A3 primary floor; buff, smoothed surface, with slip on interior
and exterior; 14, with square hollow base, from primary floor of B6, red-
slipped, burnished. Similar bowls are found at Hasanlu IIIB, also, in one case
at least, with a squared hollow base; Ziwiye (Young 1965, fig. 3, 3); Zendan
128 chapter five

(Boehmer 1961, pl. 56, 16, 17); Bastam (Kroll 1970, fig. 4, 2); Godin II (Young
1969, fig. 43, 4).
Trefoil pitchers: Figure 15, 15; buff, burnished, medium grit. Similar pitch-
ers, with or without shoulder grooves, occur at Hasanlu IIIB and A (Dyson
1965, fig. 13, lower right; Young 1965, fig. 2, 7); Bastam (Kroll 1970, fig. 1, 7; fig. 2,
6); Baba Jan I (Goff 1970, fig. 8, 1); Achaemenid Village II (Ghirshman 1954,
pl. XXXVIII, G.S. 1221d); Altintepe (Emre 1969, fig. 17, 18); Karmir Blur (Piotro-
vskii 1959, fig. 51; Piotrovskii 1969, pl. 49, 50); Igdyr (Barnett 1963, fig. 20, 2);
also at other Urartian sites; Norsuntepe (Hauptmann 1970, fig. 22, 10, middle
Iron Age).
Jar with rolled tab handles: Figure 15, 16; buff ware (Dyson 1965, 213,
fig. 13, listed in the IIIA section). Similar handles on pots occur at Hasanlu IV
(Young 1965, fig. 2, 3); Geoy Tepe A (Burton-Brown 1951, fig. 36, 102; fig. 37,
120); see also Tresors de lAncien Iran (Geneva, 1966) fig. 64, catalogue no. 672.
One-handled pitchers: Figure 16, 1 (6424), 2 (6423), both from the fill
over the floor of B10, and of Period 2; coarse ware, buff; 2 has a smoothed
broken rim suggesting it was used after breaking. Similar pitchers occur at
Hasanlu IIIB, Ziwiye, Giyan I, Sialk B (Dyson 1965, fig. 7); the neck of 1 is
missing but the general shape looks like pitchers from Bastam (Kroll 1970,
p. 73, fig. 1, 7), Altintepe (Emre 1969, fig. 17), Karmir Blur (Piotrovskii 1959,
fig. 51), and at other Urartian sites; see also Norsuntepe (Hauptmann 1970,
fig. 23, 7). At the Urartian sites this shape usually has a fine red polished
surface.
Jars with two handles: Figure 16, 3; buff, burnished, fine paste. Compara-
ble jars may be seen at Godin (Young 1969, fig. 42, 1); Nush-i-Jan (Stronach
1969, fig. 6, 9; fig. 7, 2). None of these is an exact parallel.
Pot with everted neck, simple pinched rim, grooves at shoulder:

Figure 16, 4, Figure 17 (6419); fill over the floor of B1 and belonging to
Period 2; buff orange, burnished, medium grit paste; interior of jar is scraped,
interior of neck is smoothed. Similar shaped red-slipped vessels are found
at Hasanlu IIIB (Young 1965, fig. 2, 8, 11, for shape); Ziwiye (ibid., fig. 3, 7, 13);
Zendan (Boehmer 1961, pl. 52, 2); Godin II (Young 1969, fig. 42, 15).
Small pots with plain or slightly rolled rim: Figure 16, 5 (6435),
from primary floor of D1, and belonging to Period 1; buff, burnished, slightly
130 chapter five
Figure 17. AG 6419.
hollow base; 6 (6416), fill over primary floor of B6, also Period 1; buff,
medium paste. Similar small pots occur at Bastam (Kroll 1970, fig. 5, 5); Geoy
Tepe A (Burton-Brown 1951, fig. 38, 20); Zendan (Boehmer 1961, pl. 50, 4);
Ziwiye (Young 1965, fig. 3, 7); War Kabud (Vanden Berghe 1968, fig. 29, 3, 5).
Tripod jar with hole at shoulder: Figure 16, 7 (6426), fill over floor
of B10, and belonging to Period 2; buff, smoothed surface. The neck has no
opening. Tripod vessels occur in Iran in the Bronze and Iron Ages: L. Vanden
Berghe, Archologie de lIran ancien (Leiden, 1959) pls. 111b, c, 115b, c, 145e,
146c, 165b, 167 (middle), 173a, b.
Large pot with rolled rim: Figure 18 (64-8), ht. 33.5 cm.; from primary
floor of A2, and belonging to Period 1; buff, burnished, medium grit paste.
Parallels in shape occur at Haftavan (Burney 1970, fig. 8, 3, red burnished);
Altintepe (Emre 1969, fig. 3); Igdyr (Barnett 1963, fig. 21, 4, 6,7).
Tray: one fragment of coarse ware, buff. Trays occur at Hasanlu III
(Young 1965, 75, fig. 12, apparently IIIA); Ziwiye (ibid.); Bastam (Kroll 1970,
Figure 18. AG64-8.
fig. 6, 5); Zendan (Boehmer 1961, pl. 60, 5, 16, 17, 18); S Girdan (Muscarella
1971a, fig. 30, upper left, and second from bottom); Qalatgah surface, unpub-
lished. In the fill of B7 were found two base sherds of undetermined shape
each of which has holes made before firing; they are probably not trays. A
tray from Qalatgah, referred to above, has a hole in its center; and a vessel
from Baba Jan (Goff Meade 1968, fig. 10, 26) has a hole in its center.
High-necked jars: Figure 16, 812, sherds of about five others; all are buff,
smoothed surfaces, except 10 which is buff, burnished. Similar jars occur at
Hasanlu IIIB and A; Ziwiye (Young 1965, fig. 3, 10); Bastam (Kroll 1970, fig. 3, 4,
5); Zendan (Boehmer 1961, fig. 30, 24, pl. 54, 6); Godin II (Young 1969, fig. 42,
3, 4, 6, 9, 14, 16); Pasargadae, unpublished; Altintepe (Emre 1969, fig. 2, 3, 5,
6).
Bowls or basins without handles: Figure 19, 1, coarse ware, buff; 2, buff,
lightly burnished; fragments of two others that are coarse buff, and one that
is buff, smoothed surface. Similar vessels occur at Hasanlu IIIB and A.
132 chapter five

Small pithoi, storage jars: Figure 19, 310, and fragments of others; most
are coarse, buff, and unburnished; 7 is red-slipped, 8 and 9 are buff with
smoothed surfaces, and they have grooved rims. Parallels are found at Ha-
sanlu IIIB; Ziwiye (Young 1965, fig. 4, 3); Geoy Tepe A (Burton-Brown 1951,
fig. 40, 1648); Zendan (Boehmer 1961, fig. 31, 2; pl. 54, 7); Godin II (Young 1969,
fig. 43, 1, 6, 15, 19); Bastam (Kroll 1970, fig. 5, 1, 4). Grooved rims on storage
vessels occur at these sites also.
Tab handle: Figure 19, 11; from fill in B1, belonging to Period 2; dark brown,
smoothed surface, medium grit paste; only one example found. Parallels
for these handles occur at Hasanlu IV (Young 1965, fig. 6); Hasanlu IIIB,
Ziwiye (ibid., fig. 3, 17; fig. 4, 9; fig. 10); Zendan I (Boehmer 1961, pl. 57, 811;
Boehmer 1965, fig. 75b, 74a); Godin II (Levine 1970, p. 43, drawing); cf. Geoy
Tepe A (Burton-Brown 1951, fig. 35, 284) and Nush-i-Jan (Stronach 1969, fig. 7,
2).
Horizontal handle: Buff, coarse ware, from fill in B7; only one example
found. This type of handle occurs at Hasanlu IIIA, Ziwiye, Khorvin, Giyan I
(Dyson 1965, 206, fig. 7); Godin II (Young 1965, fig. 34, 16); Baba Jan II, III
(Goff Meade 1968, fig. 10, 1214, 18; Goff 1970, fig. 7, 46; fig. 8, 1112, Period 1);
Nush-i-Jan (Stronach 1969, fig. 6, 36).
Pithoi: Many pithoi fragments were found in the fill and on the primary
floor of D1. Figure 20, 1a, b, Figure 21 (64-1), from high in the fill of A2
and apparently either Period 2 or 3a fragment of a brown buff, coarse
pithos, apparently handmade; stamped into the clay are two round sealings,
each a skidding horned creature with erect tail; features cannot be made
out but both creatures seem to be of the same species. To their right is a
stamped curved-sided square with a round depression in the center; below
the impressions are hand-impressed inverted V marks. I cannot find exact
parallels for these sealings, but one should compare sealings from Urartu
(van Loon 1966, p. 156, F11; p. 159, E15, EI6; p. 161, G2; Barnett 1959, fig. 6, 15;
see also Mallowan 1966, p. 198f., fig. 134, 7, 7th century bc). Three practically
complete buff pithoi with scraped surfaces were found on the primary
floor of D1: Figure 20, 24, Period 1; 2 and 3 have raised triangles on the
shoulder, 4 has a rope or corded design, Parallels for the triangle decoration
occur at Karmir Blur (Piotrovskii 1969, pl. 61, sunken triangles); Kayalidere
(Burney 1966, fig. 15, sunken triangles); Patnos (Ankara Museum); Kef Kalesi
(Bili, gun 1965, pl. VIII, sunken triangles, corded also). Vessels with corded
decoration on the shoulders also occur at Urartian sites, viz., Kef Kalesi, op.
134 chapter five
cit.; Barnett 1959, pl. IV. Other pithoi have coarse, buff, unburnished surfaces:
Figure 20, 6, 8; or slightly burnished buff surfaces, 7; or plain buff, smoothed
surfaces, Figure 22, 1, 3, 4; or a red-slipped buff surface, 2 (two of these
were found). The flat-ledged type, Figure 22, 2, has parallels at Hasanlu
IIIB and A; Kayalidire (Burney 1966, fig. 15, fig. 16); the other pithoi have
parallels at Hasanlu IIIB and at Urartian sites, viz., Altintepe (Emre 1969,
pl. VI).
Figure 21. Pithos sherd with seal impressions, AG 64-1. Ht. 12.5cm.

136 chapter five
Pithos: Figure 23, 2 (6442), Figure 24 (Muscarella 1971b, fig. on p. 44; Dyson
1965, fig. 13, lower left in IIIA section); found on top of and apparently within
a pith os on the floor of D1the southwest pithos shown on the plan;
deep red, well burnished, red-slipped; inside plain and uneven; triangles
and bands on surface scraped and lighter than rest of vessel; traces of
white paint on the bands and triangles; ht. 56 cm., rim diam. 54 cm., base
diam. 19cm., carination diam. 47, 5cm. A very similar, but larger vessel
was published from Patnos (van Loon 1966, fig. 3), where the triangles
are also filled with white paint; others, unpublished, are also from Patnos.
See also a similar vessel and decoration, but cruder, from Armavir Blur
(Piotrovskii 1969, pl. 69); compare for general shape a vessel from Karmir
Blur (Piotrovskii 1969, pl. 55).
Pot stand: Figure 23, 1 (6427); from the fill over the floor of B5 and belong-
ing to Period 2; burnished red-slipped ware. Rolled upper and lower rims;
four oval cutouts around the middle.
Askos: Figure 20, 5 (6434), Figure 26; from the fill over the floor of C1; red-
slipped buff ware. The vessel is egg-shaped with upright spout and handle;
handle is grooved with a clay rivet at the base, and two rivet-like impressions
over this. Askoi occur at Karmir Blur (van Loon 1966, fig. 11; Piotrovskii
1969, pl. 58, but painted). See Vanden Berghe 1968, p. 117, fig. 144, for an
example from Luristan; examples in Copenhagen and the Louvre are said
to have come from Luristan (Contenau 1935, pl. XVIII, top; M.L. Buhl, Acta
Archaeologica 21 [1950] p. 197f., fig. 46, 47). Other askoi are reported from
Patnos (Mellink 1965, p. 142). More examples are known further west: Hama
(Riis 1948, fig. 84); Cyprus (Karageorghis 1969, fig. 31, 7th century sc). Related
shapes occur at Persepolis (Schmidt 1957, pl. 71, 9, 72, 13); Achaemenid
Village (Ghirshman 1954, pl. xxxv, G.S. 1270; pl. XXXVIII, G.S. 1176). False
terracotta rivets occur at Hasanlu IIIB and A (Young 1965, fig. 2, 13); Geoy
Tepe A (Burton-Brown 1951, fig. 37, 121); Yanik Tepe (Burney 1962, pl. XLV,
fig. 30); Baba Jan I, II (Goff Meade 1968, fig. 6, 19); Zendan (Boehmer 1965,
fig. 75, c); Achaemenid Village (Ghirshman 1954, pl. XXXIX, G.S. 1249d). For
a later shape related to the askos, see Schmidt 1957, pl. 72, 13.
Nipple-base vessel: Figure 23, 3, Figure 25 (6438); also 6443; both exam-
ples from the fill of D1, burnished red-slipped buff ware. I can find no pub-
lished examples but am informed that similar vessels have been found at
Cavustepe.

138 chapter five
Figure 24. Red-slipped pithos, AG 6442.
Figure 25. Nipple-based vessel, AG 6438. Ht. 22.5cm.

Figure 26. Red-slipped askos, AG 6434. Ht, 9.5cm.
Stem and base of a goblet: From fill of D1; red-slipped burnished; light
brown interior; fine clay core; clearly the finest red burnished vessel at
Agrab. The goblet is one of the most characteristic shapes in the Urartian
repertory. They are found at Bastam (Kroll 1970, p. 73 for bibliography, fig. 1,
4); Haftavan (apud Kroll 1970, p. 73); Kayalidere (Burney 1966, pl. XV, b);
Altintepe, earlier level (Emre 1969, fig. 19); Karmir Blur (Piotrovskii 1959,
fig. 50); Toprakkale (C. Lehmann-Haupt, Armenien Einst und Jetzt [Berlin,
1931] p. 567).
From the foregoing we see that the preponderance of the pottery at Agrab
Tepe consists of buff wares, with coarse, burnished, and smoothed surface
types in approximately equal proportions; buff unburnished wares seem to
follow next in quantity. Medium and fine paste interiors occur mixed among
these groups with no correlation to surface features other than that coarse
wares do not have fine paste. In lesser quantity, but still considerable and
very noticeable, are the red-slipped wares, usually slightly or well burnished.
Aside from the unique vesselsthe askos, nipple vessels, the pot stand,
and the triangle-decorated pithoia few pithoi and small bowls are of this
fabric. Rare, but in evidence, are a few burnished gray bowls.
140 chapter five
The buff wares fired from yellowish through pinkish to orange. Often, gold
flakes (mica?) are visible on the surface. Some of the red-slipped bowls also
have these gold flakes on the surface, suggesting they were made locally,
from the same clay source as the buff wares.
This ceramic collection makes it quite clear that Agrab Tepe was an
Iron III site as defined by Dyson and Young, belonging to the late buff ware
horizon.4 Many of the sites of this period, as we have seen in referring to
pottery parallels above, have not only the characteristic buff wares, but also
the red-slipped wares; and also a small quantity of burnished gray wares viz.,
Ziwiye, Achaemenid Village, Giyan I (Young 1965, pp. 59, 66, 68), Baba Jan
(Goff Meade 1968, p. 116), Bastam (Kroll 1970, p. 70), Godin II (rare, personal
communication with T. Cuyler Young, Jr., who also informs me that there
are also a few red-slipped wares at Godin II). This configuration of pottery
is characteristic for Iron III sites and need not be elaborated on here. Future
research will have to define the cultural relationship and significance of the
occurrence of red-slipped wares at Iron III sites and at most Urartian sites.
Noticeably lacking at Agrab Tepe are the fine wares recorded at Hasan-
lu IIIB, Ziwiye, and Qalatgah (Young 1965, pp. 55, 59 ff.; Muscarella 1971b,
p. 46f.), Yanik Tepe (Burney 1962, pl. XLV) and Pasargadae, unpublished;
also the incised wares found at Ziwiye and Zendan (Boehmer 1965, fig. 75;
Boehmer 1967, fig. 9). Painted pottery is lacking at Bastam, Godin II (except
for three sherds, personal communication from T. Cuyler Young, Jr.), Geoy
Tepe A, and Zendan. What, if any at all, are the chronological implications
of this lack of local painted pottery at Agrab Tepe cannot yet be established.
Actually it may have no chronological significance; rather, it could simply
mean that it was a luxury product, and not needed at Agrab (but what of its
lack at other Iron III sites?).
Metal
Bronze helmet earflap: Figure 27, 1 (64-9), Figure 28; fill over floor of B8.
The border has two grooves and an exterior flat ledge with holes. Traces of
thread found in situ on obverse, connecting holes; traces of leather found
on both sides, on the obverse under the thread. Earflaps of the very same
shape have been found at Hasanlu IV, but with decoration (unpublished).
T.A. Madhloom, in The Chronology of Assyrian Art (London, 1970) p. 38, says
separately made earflaps began in the 7th century, but this is contradicted
by the Hasanlu evidence.
4 Young 1965, pp. 53 ff, 72 ff.; Dyson 1965, pp. 203 ff. In addition see the pottery analysis of
Kroll 1970, pp. 67 ff.

Figure 27. Metal objects from Agrab Tepe.

142 chapter five
Figure 28. Bronze earflap, AG 6449 L. 17.7cm.
Bronze trilobate arrow: Figure 27, 2 (64-6); upper fill of A2; remnants of
wood in the shaft. The distribution of these arrows is quite extensive in the
Near East and in Europe, see T. Sulimirski, Scythian Antiquities in Western
Asia, Artibus Asiae 17 (1954) p. 295f.; R.V. Nicholls, Old Smyrna: The Iron Age
Fortifications , BSA 5354 (19581959) p. 12; P.R.S. Moorey, Catalogue of
the Persian Bronzes in the Ashmolean Museum (Oxford, 1971) p. 87; Boehmer
1965, p. 773f., note 98 (n.b., Boehmer says they are found in the Phrygian level
at Gordion but surely this is an error: they are found only in post-Phrygian
levels). There is as yet no certain evidence that these arrows predate the 7th
century in Iran (Dyson 1965, p. 207).
Bronze arrow with two wings: Figure 27, 3 (6457). Similar arrows, with-
out the side hole, occur at Hasanlu IV; Achaemenid Village II (Ghirshman
1954, pl. XLIV, G.S. 2104); Karmir Blur (Piotrovskii 1959, fig. 81, right; fig. 84,
left); Gordion (R. Young 1953, p. 164f., 166, fig. 10, 6th century bc); and Smyrna
(Nicholls, Old Smyrna, 130f., pl. 6d, right, about 600bc).
Bronze arrow: Figure 27, 4 (6440); fill in D1; leaf shaped, flat on one side,
with median strip on the other; solid ridged tang; found bent from use.
Iron arrows: Figure 27, 8, 9, 11, 12, 13 (6410), from the fill of B8, and probably
from Period 2 (a total of seventeen arrows were found here); 10 (6437), from
the fill of D1; 5, 6, 7 (6451), East Trench; a) flat blade, solid tang, 57; b)
blade oval in section, solid tang, 810; c) blade oval in section, with collar
and solid tang, 11; d) blade oval in section, tang hollow but with iron core,
12, 13: traces of reed binding remain here. A total of twenty-one iron arrows
were found, many very corroded. Similar arrows occur at Hasanlu IV; Bastan
(Kleiss 1970, p. 54 f., pl. 34, 1, iron and bronze); Ziwiye, unpublished; Sialk B
(Ghirshman 1939, pls. LXXI, S892e; LXXV, 5923c, d; LXXVII, 5973a); Haftavan
(Burney 1972, pl. VIII, b); Kayalidere (Burney 1966, fig. 21, 8, 9; pl. XIII); Karmir
Blur (Barnett 1952, fig. 13); Altintepe (zg 1966, pl. XXXIV, 16); Toprakkale
(Barnett 1954, fig. 15); Igdyr (Barnett 1963, fig. 37, 47, 11); Nimrud (Mallowan
1966, 11, fig. 332, a-c).
Iron lance head: Figure 27, 14 (6431); from fill in C1; very corroded; traces
of wood in shaft.
Iron shafthole ax: Figure 27, 15 (6421); fill over primary floor of B7; very
corroded; traces of wood in shaft.
Iron tool, ferrule?, plowshare?: Figure 27, 21 (6439); from fill in D1;
hollow but for inner 9cm. of tip. A similar, but smaller, object was found in
Hasanlu IV; cf. also J. Deshayes, Les Outils de Bronze II (Paris, 1960) pl. XVI,
16, 1235; Achaemenid Village (Ghirshman 1954, pl. XLIV, G.S. 2109).
Bronze wood clamps (three): Figure 27, 16 (6414a, b, c); fill over primary
floor of B3; pieces of wood were found adhering to the inside of one example.
Bronze boss (two): Figure 27, 17 (6420); fill over floor of B10. Cf. Zendan
(Boehmer 1965, p. 773f., fig. 77a); Sialk B (Ghirshman 1939, pl. LVI, S819);
Karrnir Blur (Piotrovskii 1969, fig. 79).
Bronze stud: Figure 27, 18 (6441); fill in D1; ends bent out intentionally.
Iron knife blade: Figure 27, 19 (6430); fill in C1; cracked; flat in section at
rounded end; tapers at blade end; a rivet pierces the metal at the grip end.
144 chapter five
Figure 29. Stone, bone, and terracotta objects from Agrab Tepe.
Bronze hook (?): Figure 27, 20 (6448); from East Trench fill; thick, twisted
circular shaft, divided and flattened at both ends.
Bronze bracelet fragment: Figure 27, 22 (6447); from fill in East Trench;
animal or snake head at ends.
Bronze bracelet fragment: Figure 27, 23 (6432); fill in C1; may have
snake-head ends.
Bronze bracelet: Figure 27, 24 (6415); from primary floor of B3; probably
a childs bracelet; plain ends slightly overlap.
Bronze bracelet fragment: Figure 27, 25 (64-2); upper fill of A2; the ends
are flattened with square corners and hollow depression; arc decorated with
two rows of shallow lines.
Bone; Stone; Terracotta; Ivory

Bone implement, shuttle (?): Figure 29, 1 (6429); fill in C1; pointed at both
ends; highly polished on upper surface. Two of these were found, the second
in the East Trench fill.
Ivory fragment: Figure 29, 2 (6456); unstratified; triangular in section; no

decoration.
Bone bead: Figure 29, 3 (6449); fill in East Trench.
Two terracotta beads: Figure 29, 4 (6428); fill of C1; (6418), fill in B8;
buff, perforated for stringing. A third bead found in fill over primary floor of
B7.
Terracotta whorl or wheel fragment: Figure 29, 5 (6412); upper fill in

B area; buff.
Stone quern: Figure 29, 6 (6444); fill in D1; smooth on flat surfaces; break
may be unintentional.
Stone object: Figure 29, 7 (6446); unstratified; gray stone, rough on upper
surface, polished on lower; a projection below is broken. A pivot stone?
Whetstone: Figure 29, 8 (6413); fill over B area; very fine grained dark
brown stone; hole at one end.
It is obvious that the elaborate building at Agrab was built to serve as a for-
tified structure. The massive walls, tower-gate, and piers preclude another
interpretation. And the hoard of slingstones, the helmet earflap, and the
arrows and iron lance, reinforce this opinion; the pithoi would have served
as storage vessels for the presumed garrisons food.
On the other hand, one might conclude that a fort at Agrab makes little
sense. It was built not on a high place on one of the ridges, but in an
exposed position on the plain, in which position it could not have served as a
watchtower. Moreover, the building is relatively small and presumably could
not have contained many soldiers comfortably, even with an upper story.
Agrab is also within sight of Hasanlu and may have had some relationship
with that city; why then should a fort have been built so close?
At Bastam, Kleiss excavated an isolated building situated about 700 me-
ters from the citadel. He suggested that this building, or castle, might have
been built by an enemy force besieging Bastam.5 Such a conclusion cannot
be presently proven, nor can we make a similar interpretation with respect
5 Kleiss 1970, pp. 40 ff.

146 chapter five
to Agrab and Hasanlu. The people who built Agrab expected it to be a

permanent structure, not a temporary fort. The fact that it was destroyed
and rebuilt demonstrates its importance.
The geographical shortcomings notwithstanding, I believe the word fort
best describes Agrab. It is also possible that the same people who lived at
Hasanlu, or their allies or subjects, built Agrab. The similarity in the con-
struction of the walls of Agrab and Hasanlu IIIB, as well as the similarity of
their pottery, lend support to this suggestion. If there was indeed a relation-
ship, one might have to assume that the spring near Agrab was important
and had to be protected, although there is no indication that the spring was
protected in the earlier Iron II Period. And perhaps the rock outcrop close
to the spring suggested itself to the architects as a natural spot on which to
build a fort.6 One might also speculate that some feature or location other
than the spring, not now known, might have been in need of protection. The
fact remains that someone did build a fortified building in the low plain on
a rock outcrop next to a spring. It should be recalled that Nush-i-Jan, to the
south, was also built on a rock outcrop in the middle of a plain.
It is tempting to speculate further that the fort was built for, or at least
served, another function, namely to protect something within its walls. The
odd subdivided stone-paved room could have had a special function of
some importance unknown to us. Insufficient data, however, prevent further
consideration of this idea.
A question that must be asked, if not easily answered, is: which people
built Agrab? If the same people who lived at Hasanlu, then the question
covers that site also. Were they Urartians? Indeed, the ceramic evidence
informs us that Urartian pottery was used at Agrab, though it does not tell
us that Urartians either built or staffed the fort; the pottery could have been
imported by non-Urartians. As Kleiss has stated, no Urartian site presently
known in Iran exists in a plain;7 and no Urartian site known to me has a
plan similar to that of Agrab. Several Urartian citadel walls have a series of
uniformly spaced piers without a tower, and so, too, does Nush-i-jan;8 but
none has the unique Agrab plan.
6 van Loon 1966, p. 38, states that Urartians built fortified outposts to guard water sup-
plies.
7 Kleiss 1970, pp. 40 ff. Note that Altintepe in Urartu was built on a low hill in the plain:
zgu 1966, p. 37 f.
8 Kleiss 1970, figs. 34, 36; Kleiss, Bericht uber Zwei Erkundungsfahrten, figs. 18, 18a; van
Loon 1966, fig. 6.

What of the Medes and Manneans? Here, too, insufficient evidence pre-
vents a conclusive answer. Given the geographical problems, and recogniz-
ing that different peoples and armies must have traversed the region in the
seventh century, any suggestion becomes a mere guess. I therefore see no
alternative to leaving the question of the ethnic identity of the builders and
occupiers of Agrab open for future research. At the same time one is tempted
to suggest that the occupiers of the site could have been either an Urartian
garrison, using local help to build the fort (but who planned it?), or a local
garrison of Manneans. But, to repeat, we do not know.
Whoever the people were who lived or worked at Agrab Tepe, they used
the same basic types of pottery for their kitchen needs as that used by
the inhabitants of contemporary cities and towns in western Iran. They
also used a pithos type common to Hasanlu IIIA and to several Urartian
cities. Moreover, some of the inhabitants owned a few exotic pottery vessels,
apparently, all imported from Urartu. Therefore we may presume that they
were in familiar contact both with Iranian and Urartian cities.
The inhabitants at Agrab Tepe stored their grain in large and small pithoi,
at least one of which was impressed with seals, and they ground their grain
on the premises. Aside from one possible agricultural tool, there are no other
indications from the material remains to suggest that they were farmers.
However, we do not know what was not preserved for us to find. Their water
was obtained from the neighboring spring, and, as stated, they were within
walking and viewing distance of a fortified city, Hasanlu. They used bows
and arrows, slings, and lances as weapons, and they had body armor.
No luxury items aside from a few terracotta beads, and a few small (for
females?) bronze bracelets were recovered. Some kind of wood furniture or
apparatus was used, of which only the bronze clamps now remain.
We turn now to a discussion. of the chronology of Agrab Tepe within
the Iron III period. To begin with, it must be pointed out that specific
dates for the beginning and end of most sites of this period have yet to
be firmly established. Speaking generally for northwestern Iran, Iron III
begins sometime after the destruction of Hasanlu IV in the ninth century
bc, presumably after a hiatus of still undetermined length. But the complex
and still unresolved chronological difficulties surrounding the beginning,
flourishing, and end of the Sialk B culture to the south play a crucial role
in any discussion of the end of Iron II and the beginning of Iron III, and not
only for central Iran, but also for the north.
A brief discussion dealing with the opinions of various scholars con-
cerned with Sialk: Ghirshman and Porada see Sialk B as an Iron II site
both in culture and date, terminating about 800 bc, about the same time
148 chapter five
as Hasanlu IV.9 Dyson and Young accept the possibility that Sialk began in
the late ninth century, contemporary with the last phase of Hasanlu IV, but
see the culture continuing to exist until about 700bc (Young), or to about
650bc (Dyson).10 Goff Meade seems to agree with this, preferring Youngs
final dating to that of Dyson.11 She and Young also still use the term Iron II
to define the flourishing of Sialk, Goff Meade calling the eighth century late
Iron II, which suggests that Iron II continued to exist at Sialk for at least a
century later than in the north.12 Boehmer attempts to divide Sialk B into
an earlier and a later period (B1 and B2), the former beginning in the late
ninth century, the latter beginning about 770760bc, after the destruction
of Hasanlu IV, and ending about 690 680bc.13 The question to be answered,
considering these various opinions, is: do we consider Sialk B to be strictly an
Iron II culture, contemporary with but divergent from Hasanlu IV (Porada,
Ghirshman), or initially an Iron II culture that began in the ninth century
and continued (uninterrupted) into the eighth century (or slightly later)
(Dyson, Young, Goff Meade, Boehmer), developing into what archaeologists
call Iron III culture, and perhaps even having been the precursor of that
culture? How one interperts the nature of the anomalous Sialk B material
(only cemetery remains, let it be noted), and also perhaps the early phase
of Zendan, will determine whether one sees Sialk as Iron II, late Iron II,
or Iron II/III. Surely a chronological and cultural distinction for the terms
Iron II and III may have to be established.
Whatever the outcome of discussions concerning the culture and chro-
nology of Sialk B,14 Agrab Tepe remains an Iron III site, and to that site we
now return.
9 R. Ghirshman, The Arts of Ancient Iran (New York, 1964) p. 279; Ghirshman 1939, p. 95;
Porada 1965, p. 107.

10 Dyson 1965, pp. 207 f.; Young 1965, pp. 78, 81 f.
11 Goff Meade 1968, pp. 121, 125.
12 Goff Meade 1968, pp. 121, 125; Young 1967, p. 24, note 71. By Youngs own terminology
some of the Sialk B material (but which?) must be considered as Iron III, op, cit., pp. 27f.; and
Goff Meades late Iron II overlaps with Youngs Iron III.
13 R.M. Boehmer, Zur Datierung der Nekropole B von Sialk, AA (1965) pp. 802ff. Note
that some of the pottery in his earlier-period tombs occurs also in his later tombs. To add
to the confusion about the dates of Sialk B, note that of the five ceramic parallels Young
finds between Hasanlu IV and Sialk B (Young 1965, pp. 76f.), only one, the gray ware spouted
pitcher, is to my mind a strong parallel; and of the nineteen nonceramic parallels he cites
between the two sites, at least fifteen are in Boehmers B2 late tombs (Young 1965, p. 76,
note 28)!
14 My present opinion is that Sialk B existed into the 8th century, but I have no strong
opinion as to whether or not it can be stated that Sialk existed in the 7th century, nor if it
began to exist in the 9th century. But note that if, in fact, the designation Iron III is to be used
At Agrab Tepe five C14 samples were tested for dating.15 The results given
here use a half-life of 5730 years and should be corrected by a MASCA cor-
rection factor of +50 years (as of 1972). From the floor of C1, P-895, char-
coal: 79556bc (845 bc); from Period 1 of area D, P-980, charcoal: 667 58bc
(717 bc): ave. 781bc 57. From Period 2 fill of area D, P-979, charcoal: 581
53 bc (631bc). From high in the fill of Area A, P-893: 408 48 bc (458 bc).
From high in the fill of Area A, P-894, charcoal: 710 57bc (760 bc); and a
sample, probably from Period 2 of Area B, P-879, a burned beam: 513 56bc
(563bc): ave. 597bc55. We thus have an outer range of dates for Period 1 to
be 838724bc; for Period 2, 652542 bc; and possibly for Period 3 (assuming
P-893 to be from this time), 506410 bc P-894 is assumed to be an aberrant.
Given the chronological range of over three hundred years it seems that we
should accept these carbon dates as a guideline rather than as data pointing
to specific historical dates.
Pottery comparisons with other sites in Iran and Urartu allow us to make
better judgments about chronology. We have seen that there are ceramic
parallels between Agrab and most of the known Iron III sites, In terms of
quantity, which of course could be accidental, the strongest ties are with
Hasanlu IIIB and A, Bastam, Ziwiye, Zendan (I and II), and Godin II; other
ties, less strong, are with Haftavan, Baba Jan (I and II), the Achaemenid
Village (I and II), and Geoy Tepe A. A brief summary of the chronology of
these sites is in order here; it will be seen that most were seventh century bc
sites that ceased to exist about 600bc
Very little of the Hasanlu IIIB and A material is available for study, and
basically the two periods remain unpublished.16 What may be said at present
only or mainly as a cultural termto define the appearance of painted wares and oxidized
buff waresand not simply as a chronological term signifying a period following upon the
destruction of Hasanlu IV, then Sialk B could be called Iron III from its inception. This would
obtain even if we accept that part of Sialk was contemporary to Hasanlu IV. An important
task for archaeologists is to learn if the painted wares of Sialk influenced the triangle ware
pottery of the north, and when this influence occurred (see Goff Meade 1968, p. 125). This
problem is the more important because similar painted wares occur in Urartu, viz., Van (von
der Osten 1952), Altintepe (Emre 1969, fig. 21, 22, pl. IV, v), Karmir Blur (rare; Piotrovskii 1969,
pl. 58), and in Anatolia from the 8th century on at many sites, viz., T. zg Kltepe and its
Vicinity in the Iron Age (Ankara, 1971), figs. 8 ff., pIs. XIV, XXI, 2; T. and N. zg, Ausgrabungen
in Karahyk (Ankara, 1949) pl. XXXI, 3, XXXII, 3; G.E.S. Durbin, Iron Age Pottery from the
Province of Tokat and Sivas, Anat. Stud. 21 (1971), pp. 104 f., fig. 3, 15, 13; cf. also pottery from
Gordion and Bogazkoy.
15 R. Stuckenrath, W.R. Coe, E. Ralph, University of Pennsylvania Radiocarbon Dates IX,
Radiocarbon 8 (1966) pp. 348 f. Recent information (1972) suggests that we may have to revise
past C14 dates upward again.
16 For general remarks and bibliography on Hasanlu III, Dyson 1965, Young 1965, pp. 53f.,
57ff., 72 ff.
150 chapter five
about the two levels is this: the beginning date of IIIB is still not known;
it could be from about 750 to 700, or perhaps even later, as it is clear that
IIIB followed upon a squatters settlement over Period IV (called IVA by
Dyson) of unknown duration. The destruction of IIIB (on parts of both the
western and eastern areas of the mound ash and charcoal layers document
a fire) occurred sometime in the seventh century; this is not in doubt. On
the western part of the mound IIIB walls were partly reused in Period IIIA,
and some new walls were built over the earlier ones; on the eastern side of
the mound there is an ash and trash layer 50 cm. wide over the IIIB level.
IIIAs beginning, and more important, its termination date are still major
problems. It is possible, to my mind, that IIIA ceased to exist (abandoned?)
close to 600 bc; more excavation and analysis of pottery, however, may
make it necessary to extend this date well into the sixth century, beyond
585bc.17
The end of the settlement at Bastam has been dated by Kleiss and Kroll
to the late seventh century or early sixth century bc on the basis of Urar-
tian pottery comparisons.18 This date seems acceptable on the evidence pre-
sented (and neatly ties in Bastams destruction with that of Agrabs [Period 1,
at least], especially since both sites depend a great deal on Urartian remains
for their chronology).
As Dyson has stated, any discussion of Ziwiye must distinguish the
archaeological site itself from the so-called Ziwiye treasure.19 He has pro-
posed a dating of about 750 for the beginning of the site and a terminal date
of about 600 for its abandonment.20 Young and Boehmer generally agree
with this range of dates.21 There seems little doubt but that the final period
at Ziwiye occurred either in the seventh century, probably toward the end
of that century, or possibly early in the sixth century.
17 Dyson 1965, pp. 211 f., and Young 1965, pp. 81 f., have IIIA continue into the Achaemenid
period. Kroll 1970, p. 76, note 105, suggests that Hasanlu IIIA ended ca. 600 bc on the basis of
the triangle-pithoi found at Hasanlu and Urartian sites. If C14 dates are to be pushed back in
time, this situation would support an earlier date for the termination of Hasanlu IIIA than
suggested by Young and Dyson. For a hiatus between Hasanlu IV and III, Young 1965, pp. 57ff.,
80.
18 Kleiss 1970, p. 57 f., accepting the possibility for an 8th-century beginning date; Kroll
1970, p. 76.
19 Dyson 1965, p. 206.
20 Dyson 1965 and R.H. Dyson, Jr., Archaeological Scrap Glimpses of History at Ziwiye,
Expedition 5, 3 (1963), pp. 35 ff.

21 Young 1965, p. 82, correctly allowing for an early 6th-century termination; Dyson 1965,
p. 206.
Zendan had two settlements. The beginning of the earlier one, period 1, is
dated by Boehmer close to 800bc, on the basis of parallels with Hasanlu IV;
he dates the end of the second settlement, II, to the late seventh century.22
Young dates the beginning of the earlier period later than Boehmer, pre-
ferring a date between 750 and 650 bc, but he also believes that the site
continued to the end of the seventh century (for the second period), being
contemporary with the end of Ziwiye. Thus he agrees with Boehmer that the
late seventh century was the final date at Zendan; Dyson has also supported
this dating.23 The strong parallels between Zendan II and Ziwiye pointed out
by Young make it clear that about 600 bc is the probable date for the end of
period II.
The excavator of Godin Tepe, T. Cuyler Young, Jr., has cautiously given a
wide range of almost 200 years, 750 to 550 bc, as the time within which the
columned hall and fortress were built.24 It would seem from the pottery evi-
dence, however, that the end of Godin II could be placed in the late seventh
or early sixth centuries, given the parallels with Ziwiye and Zendan II (not
to mention Agrab, to avoid a circular argument). Nor do I think it can be
demonstrated on the evidence available that Godin II was built much before
the beginning of the seventh century bc.
Geoy Tepe A is a mixed complex, and it cannot help in dating any Iron III
site; rather, it must be dated by comparisons with other sites. Nor can the
limited remains from the upper levels at Haftavan at present be of help to
us in chronology.25 It would seem that the levels could be dated only from
evidence available at other sites.
Goff Meade has compared the ceramics from Baba Jan to those found
at Pasargadae, Godin II, and Nush-i-Jan.26 She suggests that Baba Jan II is
eighth century, not earlier, and that Baba Jan I is probably sixth century bc,
because of parallels with Pasargadae (unpublished). The Agrab parallels
with Baba Jan seem to be with both periods, but aside from a general Iron III
relationship, we get little specific chronological help from this site.
22 Boehmer 1961, p. 82; Boehmer 1965, pp. 736, 740, 763f.; Boehmer 1967, pp. 576, 579.
23 Young 1965, p. 82; Young 1967, p. 271; Young 1969, p. 50; Dyson 1965, pp. 201f., 211, agrees
that the early period is Iron II.
24 Young 1969, pp. 31 f.; see also Goff 1970, pp. 155.
25 Burney 1970, p. 182, suggests a late 8th- to early 7th-century date; Burney 1972, p. 142,
suggests that Haftavan was destroyed by Sargon II in 714bc, but presents no objective evi-
dence.
26 Goff 1970, p. 155; Goff Meade 1968 pp. 121 f.; actually, little of the pottery has been
published.
152 chapter five
The pottery parallels between Agrab and the Achaemenid Village are
basically in the levels I and II. These are dated by Ghirshman to the seventh-
sixth and sixth-fifth centuries bc, and he has been supported in general by
other scholars.27 None of the shape-parallels from the Achaemenid Village,
except the lip spout of level 1, are crucial enough to basically affect the
chronological relationship to Agrab Tepe.
Two other Iranian Iron III sites with a few parallels to Agrab Tepe are
Nush-i-Jan and Yanik Tepe. Both have been dated by their excavators to a
time in the seventh century bc.28
Turning now to the Urartian sites outside of Iran that have ceramic
parallels with Agrab, we find that the strongest ties, not necessarily with
respect to quantity, but to a very characteristic shape, are with Karmir Blur,
Altintepe, Norsuntepe, Patnos, avustepe, and Igdyr; lesser ties are with
Toprakkale, Kayalidere, and Kef Kalesi. Some of these Urartian sites are not
yet completely published so that we are not always able to discuss Urartian
pottery types, nor to discuss their chronology with certainty.
Although there is continued discussion concerning the precise time
when Karmir Blur was destroyed, it seems that the event must have occurred
close to 600bc, apparently before 585.29 The finds from the final, second,
period at Altintepe also seem to date to this time; and it appears that the
material from the end of the first period may also be dated within the sev-
enth century, although the site may have been built in the eighth century.30
Kayalidere has not been more closely dated than to the eighth-seventh
century. But if we can use the pithoi decorated with triangles as a guide, it
could be that the destruction here occurred around 600bc, close to that of
Karmir Blurs destruction.31
At Patnos inscriptions of several Urartian kings have been found; they
date from the late ninth through the middle eighth century bc. Later archae-
ological material is in evidence, however, and it seems clear that the site
existed through the seventh century, possibly even a little later.32
27 Ghirshman 1954, p. 20; Dyson 1965, pp. 205, 211; Young 1965, pp. 79, 82; Goff Meade 1968,
p. 125.
28 Stronach 1969, p. 16, Burney 1962, pp. 147 ff.
29 Muscarella 1965, p. 237, notes 3436; Piotrovskii 1969, pp. 198f.
30 zg: 1966, pp. 38, 46; Emre 1969, p. 291 f.
31 Burney 1966, pp. 55 ff., 79. Burney links the destruction to the Kimmerians, who first
appear in Urartu in the last years of the 8th century. Note that triangle-pithoi occur at the
termination of Hasanlu IIIA and Agrab I, events no doubt close in time but not necessarily
simultaneous; see my note 17.
32 K. Balkan, Ein Urartische Tempel auf Anzavurtepe bei Patnos , Anadolu 5 (1960) pp.
105 ff.; Mellink 1965, p. 142.

Cavustepe was apparently built in the mid-eighth century bc and contin-

ued to flourish for some time. Evidence for this comes in the form of fibulae
and Scythian arrowheads, that is, trilobate and spiked types, found in the
destruction fill.33
Toprakkale was built in the eighth century and continued to flourish until
the late seventh, judging from inscribed material found there.34 Kef Kalesi is
another site that flourished during the seventh century, as evidenced from
inscriptions, and it too may not have lasted beyond about 600 bc.35
Norsuntepe, a site to the west of Urartu proper, and probably not an Urar-
tian site, surprisingly has yielded some good ceramic parallels to Urartian
types and also to pottery from Agrab. Hauptmann distinguishes an early and
a middle Iron Age period, the earlier of which he dates about 800 bc (too
early?), the latter to the eighth and seventh centuries bc. Most of the Agrab
parallels are in the middle period.36
The preponderance of the ceramic evidence presented strongly suggests
that Agrab Periods 1 and 2 flourished during the seventh century bc; the C14
evidence generally supports this conclusion. Most of the ceramic parallels
in Iran and Urartu occur at sites dated to the seventh century. Aside from
the many seventh-century Iron III Iranian ceramic comparisons made, the
best pieces of evidence for the suggested dating of Agrab are the Urartian
vessels: the pithoi with triangles, the large red ware pithos decorated with
triangles, the red ware nipple-based vessels, the red ware askos, and the red
ware stemmed goblet, a classic Urartian shape found at many Urartian sites.
If it can be accepted as a historical fact that the Urartian cities referred to
in this paper were in fact destroyed close in time to each other, say between
the last years of the seventh century and 585 bc, we may then feel secure that
Agrab Tepe also ceased to exist during this time.
The political and archaeological history of western and northwestern Iran
in the seventh century bc is still not fully understood. Several unresolved
problems persist. First, there is the major problem of ancient geography: we
are still unable to link up satisfactorily specific areas of northwestern Iran to
the historical states and peoples mentioned in ancient texts. Consequently,
33 Mellink 1966, p. 151.

34 van Loon 1966, p. 50; A. Erzen, Untersuchungen in der Urartaischen Stadt Toprakkale
, AA (1962) pp. 294, 406 f.
35 gn 1967, p. 499.
36 Hauptmann 1970, pp. 64, 67, 73. Hauptmann, p. 71, notes that a fibula was found in the
middle period; it is a type that cannot be earlier than the late 8th century and continued to
be used for centuries.
154 chapter five
excavated sites must continue to maintain their modern names. A perusal of

the published opinions of several scholars who have discussed the ancient
position of Parsua, Mada, and Mannea from the ninth century onward
demonstrates to the archaeologist the danger of assigning an ancient name
to a modern area.37 We are not able at present to relate sites to ancient states
and then to tie these into historical events related in the texts.
Secondly, there is the problem concerned with understanding and rec-
ognizing the movements of peoples and the actions of their armies in the
area. What political event, and what army, destroyed Agrab and neighboring
sites? We know that Scythians were somewhere in western and northwest-
ern Iran after about 700 bc, as allies first of the Mannaeans and then of the
Assyrians, and that they were subsequently expelled. The date of the Median
revolt against them and their expulsion preceded by their twenty-eight-year
hegemony over western Iran, is still being debated. We are, however, able to
state that all this occurred by 585 bc, by which time the Medes controlled
all of western Iran and Urartu, and Anatolia up to the Halys River.38 We also
know that in the seventh century the Assyrians penetrated into Median and
Mannaean territory several times, although we are not sure how far north
they traveled. There were also local wars and disruptions that surely resulted
in destructions and rebuildings. In addition to these events, we now know
from the Agrab excavations that contact with Urartu existed in the south
Urmia area up tothe time of the destruction of the Urartian state. If we could
be certain that Urartians themselves were in the area around 600bc, then
another element would have to be introduced into the already confused
historical scene. Actually, as already discussed, all we can determine with
respect to the Agrab evidence is an archaeological presence of Urartians;
we cannot be certain that Urartians themselves was there. In any event,
Agrab supplies important evidence concerning Urartian material in the Sol-
duz Valley about 600bc.39
37 R. M, Boehmer, Volkstum and Stdte der Manner, Deutsch. Arch. Inst. Abt, Baghdad
3 (1964) pp. 11 ff.; also his Zur Lage von Parsua in 9. Jahrhundert vor Christus, BJV 5 (1965)
pp. 187 ff.; Young 1967, p. 14 f.; Kleiss, Zur Ausbreitung Urartus, pp. 130ff.; Louis D. Levine,
Contributions to the Historical Geography of the Zagros in the Neo-Assyrian Period, Ph.D.
dissertation, University of Pennsylvania, 1969. One also wonders if the Assyrians might not
have called the area around the south and west shores of Lake Urmia Urartu because of
their control over that area, Muscarella 1971b, p. 49.
38 Dyson 1965, pp. 208 ff.; Young 1967, pp. 12 ff.; Stronach 1969, pp. 5ff.; van Loon 1966, p. 23;
Porada 1965, pp. 123, 137 f.; Piotrovskii 1969, p. 198 f.

39 The latest Urartian inscriptions from the general area are 7th century, from Bastam and
from east of Lake Urmia; Kleiss, Zur Ausbreitung Urartus, pp. 124f., 127ff.
Given the military presence of different peoples in western and north-

western Iran, are we able to relate chronologically and historically the de-
structions at Agrab to those of Hasanlu III and to the other Iron III sites in
Iran? (Actually, some Iron III sites were destroyedHasanlu IIIB, Bastam,
Haftavan, Zendan II, Baba Jan IIBwhile othersHasanlu IIIA, Ziwiye,
Godin II, Nush-i-Janwere abandoned.)
Of particular interest for future research is the possible correlations of
Agrabs destructions to that of Hasanlu IIIB and the abandonment of IIIA.
We have seen that there are strong ties, reflected ceramically, between Agrab
and Hasanlu IIIB and A. An interesting one is the pithos decorated with
triangles, which occurs in the destruction of Agrab I and Hasanlu IIIA.
This by itself cannot make these two periods contemporary, one to one,
as there are many Hasanlu IIIB parallels also to be considered. There are
simply too many unknowns at present to allow any neater and more explicit
equation than one stating that Agrab was contemporary to Hasanlu IIIB and
A (at least in part). Whether Agrab was originally built at the same time as
Hasanlu IIIB, or slightly earlier, or later, and whether occupation continued
at Agrab after the destruction of IIIB and before the building of IIIA, and
whether IIIA continued to exist after the end of Agrab, are questions that
arise in ones mind, but to which there are no immediate answers. Perhaps
the publication of the complete Hasanlu material will shed light on these
questions.
With respect to the issue of relating the destruction of Agrab Tepe (Peri-
ods 1 and 2) with the destructions and abandonments of other Iranian
Iron III sites, two hypotheses come to mind. The first is that they all occurred
at about the same time. An event or related events occurred in Iran about
600bc, causing the destructions of Agrab and the end of the settlements
at Bastam, Haftavan, Zendan II, Ziwiye, Godin II, Nush-i-Jan, and probably
also Baba Jan II, not to mention again Hasanlu III. The activities causing
these destructions could have taken place over a period of several, say one
to fifteen, years; nevertheless, they were related. The time period covered
would extend from about 600 to 585bc. Moreover, and important, is the fact
that, given the chronological connection between Agrab Tepe and Urartu
already discussed, it could legitimately be added here that the same histor-
ical event or events may have caused the destruction of the Urartian state.
This hypothesis will obviously deserve more scrutiny, but if the chronologi-
cal link of the destruction of the Iranian Iron III sites and the Urartian cities,
based on pottery parallels, holds up, such a conclusion is not rash.
What then can be said about this alleged historical event? A date of
about 600 bc automatically eliminates the Assyrians. And there seem to be
156 chapter five
only two historical events, themselves related, that can be brought forth
for consideration. One is the exodus of the Scythians from Iran, the other
is the northward expansion of the Medes, through north Iran, Urartu, and
eventually west to Anatolia. Again, this suggestion deserves further study,
but no other large-scale action occurred in Urartu and Iran at this particular
period.
The second hypothesis, also to my mind viable in that it does not abuse
the limited evidence, is that the sites in question were destroyed or aban-
doned over a slightly longer period of time than suggested by the first
hypothesis. Thus, one could assume that some of the sites could have ceased
to exist about 600585 bc, while others could have ended about 550, say at
the time of the Achaemenid revolt against the Medes. One would then have
a time differential of twenty-five to forty years between the end of one par-
ticular site and another. In this context it must be understood that we do
not yet have an idea of what early sixth-century bc and early Achaemenid
pottery looked like, and it is quite possible that there was no major ceramic
change between about 600585 and about 550bc. The fact that pottery
analysis at our present state of knowledge might not allow us to detect a
chronological difference between pots used at different neighboring sites
over a period of a few decades is the crucial factor here. This hypothesis,
incidentally, might also cover the problem of the difference between the
destructions at Agrab 1 and 2, and Hasanlu IIIB and A, but it would be pre-
mature to push this idea now.
In any event, these are nothing more than working hypotheses, to be
challenged or supported as more ceramic, archaeological, and historical
information comes forth. The end of the seventh and the first half of the
sixth century bc in Iran and Urartu was a time of chaos, destruction, and
abandonment for its people, and it is a time of chaos for modern historians.
Bibliography
Barnett 1952
R.D. Barnett, Russian Excavations in Armenia, Iraq 14 (1952) pp. 132147.
Barnett 1954
Excavations of the British Museum at Toprakkale, Addenda, Iraq 16 (1954) pp. 322.
Barnett 1959
Further Excavations in Armenia (19491953), Iraq 21 (1959) pp. 119.
Barnett 1963
The Urartian Cemetery at Igdyr, Anatolian Studies XIII (1963) pp. 153198.
Bili, gn 1965
C. Bili, B. gn, Second Season Excavation at Kef Kalesi of Adilcevaz, Anadolu 9
(1965) pp. 1128.
Boehmer 1961
R.M. Boehmer, Die Keramikfunde vom Zindan-i-Suleiman, Teheraner Forsch. 1
(1961) pp. 8286.
Boehmer 1965
Takht-i-Suleiman und Zendan-i-Suleiman, Grabungsbericht 1963 und 1964, AA
(1965) pp. 746787.
Boehmer 1967
Forschungen an Zendan-i-Suleiman in Persisch-Azerbaidschen 19581964, AA
(1967) pp. 573585.
Burney 1962
C.A. Burney, Excavations at Yanik Tepe, Azerbaijan, 1961, Second Preliminary Re-
port, Iraq 24 (1962) pp. 134152.
Burney 1966
A First Season of Excavation at the Urartian Citadel of Kayalidere, Anat. Stud. 13
(1966) pp. 55111.
Burney 1970
Excavations at Haftavan Tepe 1968, Iran 8 (1970) pp. 157171.
Burney 1972
Excavations at Haftavan Tepe 1969, Second Preliminary Report, Iran 10 (1972)
pp. 127142.
Burton-Brown 1951
T. Burton-Brown, Excavations in Azerbaijan, 1948 (London, 1951).
Contenau 1935
G. Contenau, Fouilles de Tepe Ciyan (Paris, 1935).
Dyson 1965
R.H. Dyson, Jr., Problems of Protohistoric Iran as Seen from Hasanlu, JNES 24, 3
(1965) pp. 193217.
Emre 1969
K. Emre, The Urartian Pottery from Altintepe, Belleten 33 (1969) pp. 291301.
Ghirshman 1939
R. Ghirshman, Fouilles de Sialk (Paris, 1939).
Ghirshman 1954
Village perse-achemenidc, in Mem. de la Mission archeologique en Iran XXXVI
(Paris, 1954).
Ghirshman 1970
Masjid-i-Suleiman, Iran 8 (1970) pp. 183185.
158 chapter five
Goff 1970
C. Goff, Excavations at Baba Jan, 1968, Iran 8 (1970) pp. 141156.
Goff Meade 1968
C. Goff Meade, Luristan in the First Half of the First Millennium B.C., Iran 6 (1968)
pp. 105134.
Hauptmann 1970
H. Hauptmann, Norsuntepe , Istanbul. Mitt. 1920 (19691970) pp. 2378.
Karageorghis 1962
V. Karageorghis, Excavations in the Necropolis of Salamis 1 (Nicosia, 1962).
Karageorghis 1969
Salamis (New York, 1969).
Kleiss 1970
W. Kleiss, Ausgrabungen in der Urartische Festung Bastam, AMI 3 (1970) pp. 765.
Kpcke 1967
G. Kpcke, Neue Holzfunde von Samos, Ath. Mitt. 82 (1967) pp. 100148.
Kroll 1970
S. Kroll, Die Keramik aus der Ausgrabung Bastam 1969, AMI3 (1970) pp. 6792.
Levine 1970
L.D. Levine, Of Medes and Media, Rotunda 3 (1970) pp. 3644.
Mallowan 1966
M. Mallowan, Nimrud and its Remains I, II (New York, 1966).
Mellink 1965
M. Mellink, Archaeology in Asia Minor, AJA 69 (1965) pp. 133149.
Mellink 1966
Archaeology in Asia Minor, AJA 70 (1966) pp. 139159.
Muscarella 1965
O. White Muscarella, A Fibula from Hasanlu, AJA 69 (1965) pp. 233240.
Muscarella 1971a
The Tumuli at S Girdan: Second Report, MMA Jour. 4 (1971) pp. 528.
Muscarella 1971b
Qalatgah: An Urartian Site in Northwestern Iran, Expedition 13, 34 (1971) pp. 44
49.
gn 1967
B. gn, Die Ausgrabungen von Kef Kalesi , AA (1967) pp. 481503.
zg 1966
T. zg, Aliintepe I (Ankara, 1966).
Piotrovskii 1959
B.B. Piotrovskii, Vanskoe Urartu (Moscow, 1959).
Piotrovskii 1969
The Ancient Civilization of Urartu (New York, 1969).
Porada 1965
E. Porada, The Art of Ancient Iran (New York, 1965).
Riis 1948
P.J. Riis, Hama, Les Cimetires Crmation (Copenhagen, 1948).
Schmidt 1957
E. Schmidt, Persepolis II (Chicago, 1957).
Stronach 1969
D. Stronach, Excavations at Tepe Nush-i-Jan, 1967, Iran 7 (1969) pp. 119.
Vanden Berghe 1967
L. Vanden Berghe, La Ncropole de War Kabud, Archeologia 18 (1967) pp. 4967.
Vanden Berghe 1968
Het Archeologisch onderzoek naar de Bronscultur van Lurisian , (Brussels, 1968).
van Loon 1966
M. van Loon, Urartian Art (Istanbul, 1966).
von der Osten 1952
H.H. von der Osten, Die Urartaische Topferei aus Van , Orientalia 21 (1952)
pp. 307328.
R. Young 1953
R.S. Young, Making History at Gordion, Archaeology 6, 3 (1953) pp. 159166.
R. Young 1962
The 1961 Campaign at Gordian, AJA 66 (1962) pp. 153168.
Young 1965
T. Cuyler Young, Jr., A Comparative Ceramic Chronology for Western Iran, 1500
500BC, Iran 3 (1965) pp. 5385.
Young 1967
The Iranian Migration into the Zagros, Iran 5 (1967) pp. 1134.
Young 1969
Excavations at Godin Tepe (Toronto, 1969).
chapter six
THE IRON AGE AT DINKHA TEPE, IRAN*
To the memory of Rodney S. Young, 19071974
In 1966, The University Museum of the University of Pennsylvania and The

Metropolitan Museum of Art, working together as the Hasanlu Project,
began excavations at Dinkha Tepe, a site in the Ushnu valley, near Lake
Rezaiyeh (Urmia), in northwestern Iran. Background information about the
site and the reasons for excavating there, as well as a preliminary report
on the 1966 season, were presented in 1968 (Muscarella 1968, pp. 187196).
The reader is referred to that report to avoid repetition of the information
here.1 In 1968 a second campaign was conducted at Dinkha Tepe. The field
work was mainly concerned with Bronze Age remains, but part of an Iron II
structure was excavated and is discussed below.
It will be recalled from the earlier report that an Iron Age cemetery,
containing burials of both the Iron I and II periods, was discovered, that the
cemetery overlay Bronze Age strata, and that there were no architectural
or burial remains of the Iron III period. A terminology for the levels was
established in which the Iron II period was called Dinkha II (counting from
the top down; Dinkha I was the Islamic period), the Iron I period, Dinkha III,
and the Bronze Age strata, Dinkha IV.

1 The staff for the 1966 campaign is listed in Muscarella 1968, p. 187. In 1968 the staff
consisted of the writer and Robert H. Dyson, Jr., as Co-Directors, Christopher Hamlin, Carol
Hamlin, Matthew Stolper, Elizabeth Stone, William Sumner, and Harvey Weiss as site super-
visors, and Marie Sherman Parsons as Registrar. Most of the drawings were made by Mary
Voigt and Maude de Schauensee (1966). John Alden and Elizabeth Hopkins inked the draw-
ings; their expenses were paid for by a generous grant from the Schimmel Foundation. I wish
to thank all the individuals mentioned as well as the Schimmel Foundation for their coopera-
tion in the production of this report. I also wish to thank Robert H. Dyson, Jr., Louis D. Levine,
and T. Cuyler Young, Jr., for discussions and opinions exchanged over the years about Iron
Age problems, and for reading this report in manuscript. Of course, I alone assume respon-
sibility for the format and the conclusions expressed, and for not always following their
advice.
162 chapter six
Figure 1. Plan of excavation trenches.
This paper first reports on the Dinkha III cemetery, its burials and their
contents, and its relations with contemporary sites. Following this is a report
and discussion on the Dinkha II architecture and burials. No attempt is
made here to write a history of the Iron Age or a definitive summary of
that period. Not enough information is available at present and several
good summaries already exist (Dyson 1964a, pp. 3440; 1965, pp. 195213;
1968a, pp. 2932; Young 1965, pp. 5559, 6268, 7083; 1967, pp. 2229; Bur-
ney, Lang 1972, pp. 113126). Rather, the emphasis here is on Dinkha Tepe
itself.
The mound was first divided into large grid-squares one hundred meters
to a side, and these were then subdivided when necessary into ten-meter
excavation squares. In addition to these squares, shorter test trenches and
wells were opened at various parts of the mound (Figure I; Stein 1940, p. 369,
fig. 23 for a contour plan).
the iron age at dinkha tepe, iran 163
Dinkha III Period
One hundred and five burials were excavated on the mound. Thirty-three
of these are of the Dinkha III or Iron Age I period (Muscarella 1968, p. 189,
incorrectly listed twenty-six; see Table I). The majority of the burials were
excavated in the four northern excavation squares, the main cemetery area
of the mound, but some were found in the south and west. Whether these
latter burials were originally thinly scattered away from the main center at
the north, or whether they were part of a regular cemetery area encircling
the mound is not known, as extensive excavations were not conducted
in these areas. If there was a settlement on the mound that belonged to
Dinka III it could have been in the center and eastern sections, but no
architectural remains attributable to this period were recognized.
The dead were buried in individual graves with no markers; the brick
tombs generally opened to the east. Men, women,2 and children were buried
in the same area and apparently given the same burial rites. All the burials
of Periods III and II were placed within pits, which were then refilled; in a
few cases we were able to recognize the pit lines (Muscarella 1968, p. 190,
fig. 7).
Dinkha III burials were recognized primarily by the associated grave
goods, artifacts quite familiar to us from the Hasanlu excavations. In general,
the burials were stratigraphically lower in the fill than the later Dinkha II
burials; in some cases they were in the same stratum or were only slightly
lower than the later burials. Some Dinkha III burials were recognized as
being lower in the fill than others of the same period and these might be
earlyalthough the possibility exists that some pits were dug deeper than
others (but compare TT VII, below). In a few cases the pottery types of these
deep burials seem to support a conclusion for a suggested earlier deposition
(see below).
Twenty-three of the burials were simple inhumations while ten were
associated with built brick tombs. Of the latter, four consisted of a horizontal
row of mud bricks to one side of which was placed the body. Three tombs
(B9a, 22, B9b, 11, B10b, 13) consisted of a horizontal row of mud bricks
with a projection or arm at each end, forming a three-sided tomb that
2 No professional physical anthropologist examined the bones when they were exca-
vated, and therefore it is not certain that the sexing was always accurate. The bones are
currently being studied by Ted A. Rathbun of the University of South Carolina. For contem-
porary skeletal material see Rathbuns A Study of the Physical Characteristics of the Ancient
Inhabitants of Hasanlu, Field Research Projects (Coconut Grove, Miami, Florida, 1972).
164 chapter six
Figure 2. Burials 11, 12, 16, 18, 20, and 21 are of Period III.
enclosed the body and goods; one of these tombs (B9a, 22) had a mudbrick
floor. One tomb had an offset at each corner of the arms (Figure 2:21, partly
excavated; and Muscarella 1968, fig. 2), a feature common in the next period;
two tombs were disturbed. The main horizontal wall had two to four courses
and the arms two to three courses, the latter lower than the former. The top
course of the horizontal wall overlapped the grave area, often dug deeper
than the lower level of bricks, and in a few cases collapsed onto the body.
The bodies were oriented N-S or E-W, the former in the majority, and
although heads faced all points of the compass, those facing E predomi-
nated. The body was placed on the back or side; legs were flexed, with three
exceptions; arms were usually flexed before the face, chest, or pelvis, or
placed at the sides. Eight skeletons had one arm flexed across the body while
the other was bent back tightly, touching its own shoulder. No relationship
with regard to age or sex seems to exist in orientation or body position (for
details, Table I).
The characteristic ceramic objects of Dinkha III are the bridgeless
spouted pouring vessel, never with a handle; the pedestal-base goblet with
a vertical loop handle; and a flaring-sided bowl, either with a raised crescent
on the interior surfacecolloquially called worm bowlssometimes
with holes for suspension, or the same type bowl but without the crescent.
These types of vessels are classic diagnostic objects from the Iron I period.
Seventeen of the thirty-three burials did not contain a spouted vessel, but
nine of these had either the worm bowl or goblet; the eight others, contain-
ing only one or two vessels, were low enough in the fill to allow for a Period III
designation. Not a single burial contained all three of the diagnostic vessels
together (compare below, Geoy Tepe and Hajji Firuz).
Other Dinkha III shapes include deep carinated bowls, carinated jars with
relatively large mouths, and basket-handled teapots. These shapes continue
into Period II and by themselves are not easily distinguished into Iron I or
II.
Eighty-one vessels were recorded from the Dinkha III period, eighty from
the burials, one from the fill (Muscarella 1968, p. 193, fig. 17, left). Of these,
fifty-six were gray, twenty-three buff,3 and one was painted: thus the per-
centage of gray to buff is 71 percent to 29 percent. Whether the surfaces
were intentionally fired to these colors by controlling the oxygen within the
kilns, or whether the colors resulted fortuitously from firing to firing, or even
from uneven control in a given firing, is not clear. However, the fact that in
3 Buff is a term used for the oxidizing firing that produced non-gray (reduced) surfaces.
The colors of the buff pottery at Dinkha range from buff to light orange, orange, reddish-
orange, and red. The problem is not significant if one realizes that the Dinkha kilns produced
both reduced grays and oxidized buff colors. In the text I use the word buff in a general sense,
for the non-gray pottery, and in those particular instances where no specific color other
than buff was registered. Surfaces are categorized macroscopically as matt (A): no luster;
smoothed (B): a slight luster, with some stroke marks visible; burnished (C): stroke marks
quite visible and a definite luster. These divisions grade into one another. Interior paste is
categorized macroscopically as I: smallgrit inclusions of sand size; II: grit inclusions smaller
than sand to no grit inclusions visible; and III: coarse, with grit inclusions larger than sand
and visible. For convenience I use the abbreviated forms, e.g., IA, IIB, when describing a vessel
in the text. Next to each field number referred to in the text is a letter that gives the present
location of the object: M: Metropolitan Museum of Art; P: University Museum, Philadelphia;
T: Teheran Museum; D: discarded in the field.
166 chapter six
Period II buff pottery predominates might suggest that the coloring was con-
trolled (Young 1965, p. 55).
Within the gray pottery repertory, burnished surfaces outnumbered
smoothed surfaces more than two to one; only one had a matt surface.
Concerning the buff pottery, of which orange predominated, twelve were
smoothed, two burnished, eight matt, and one was red-slipped. About a half-
dozen vessels, gray and buff, had traces of mica flakes on the surface. Most
of the vessels were made of a paste that had no visible inclusions; about a
half-dozen had medium-sized grit, and only one (a worm bowl) was made
of coarse ware. Thus, the vessels may be categorized as neither of fine nor
of coarse ware, but rather what has been called common ware (Young 1965,
p. 55). Note that gray and red-slipped wares, and vessels with mica flakes,
occur in small amounts in the preceding Dinkha IV period.
The number of vessels placed within a burial varied from one to four,
and there seems to be no connection between the number of vessels, or,
indeed of burial goods in general, to inhumation or brick tomb, or to age
and sex (Table I). Some of the vessels, including all types, had obviously been
damaged in antiquity. But this fact did not deter their inclusion in a burial
and suggests that vessels placed in a burial were the same ones normally
used in the contemporary households.
Four burials contained weapons; thirty contained jewelry, worn by men,
women, and children, indicating that the dead were adorned as well as
clothed. In no burial of the Dinkha III period was an iron object found.4
An exception could be B10b, 11, discussed below, and which I consider to
belong to the Dinkha II period. Only one burial contained gold, B9a, 26,
and only one burial (B9a, 23) a cylinder seal (Table I).
Food remains in the form of sheep/goat bones were found in only three
burials, but it is quite possible that boneless meat and even liquids, all now
disappeared, were placed in some burials (see B9a, 15).
As stated above, it was possible to recognize that a few burials were
deposited earlier in time than others. To these examples we now turn. In
TT VII, Grid L, a square 2.5 1.5m, two period III burials were discovered, and
by a stroke of luck one had been deposited directly over the other (Figure 3).
1, the later, found in stratum 2, was an inhumation of a young adult female
placed in an extended position on the R side, oriented N-S, head to N. The
4 When a metal object is mentioned in the Dinkha III section it is bronze (not analyzed),
except for the gold earrings; and when a spouted vessel or goblet is mentioned, it is a
bridgeless spout and a pedestal-base goblet. In the Dinkha II section a spouted vessel always
means a bridged one.
Figure 3.
168 chapter six
Figure 4. Test trench VII, burial 2.
body had a pin, plain loop rings, bracelets, and a torque,5 (all corroded, so
discarded,) a spouted vessel (933P, like 234 in Figure 16), a bowl with two
holes (882D), and a carinated jar (938T), all gray IIC ware.
In stratum 6, but apparently cut from 4, was found 2 (Figures 3 and 4).
This was a young adult male in an extended position with the legs slightly
flexed, E-W, head to E, placed in a brick tomb (whether the tomb had arms or
not we do not know). The skeleton had a plain penannular bracelet on the
R wrist (316P), and a tanged dagger, with wood remains on the tang and a
wood peg still in the tang hole (1000P), placed behind the head. In the same
position was a red-orange IIB spouted vessel with a missing tip (237P), and
a tall gray IIB goblet (229P). No other goblet found at Dinkha has the same
shape, with straight walls, nor does any other spouted vessel have the same
5 By torque I mean a penannular necklace, at Dinkha made from one piece of metal,
and not necessarily twisted. In two burials, both of Period III, B9a, 22, B9b, 16, originally
penannular necklaces had their ends tied together, I have not considered them as true
torques.
Figure 5. Painted vessel from B9a, burial 25.
body structure, with relatively tall and straight inner vertical section of the
spoutexcept a vessel from B9a, 24 (Figure 6, 936) also early. The vessel
also lacks a beard below the spout, and has a narrow mouth. Doubtless this
is one of the earliest Iron Age burials at Dinkha.
In B9a a number of period IIIbut no IIburials were excavated in
strata 5a and 6; six burials were also found below these in stratum 7, cut
into the Bronze Age deposit. These latter could be early period III burials.
One of these burials stands out from the rest because it contained the only
painted vessel from the Iron Age at Dinkha.
B9a, 25 contained the inhumation of a mature adult male flexed and
placed on his back, E-W, head to E. He wore a toggle pin with a finely
decorated top (Figure 3, 473P), and a necklace of paste beads (1006T); a gray
IIC goblet (696T) and a polychrome jar (420T) were the other grave goods.
The latter has a cream surface overpainted on the upper body with reddish
brown hatched triangles outlined with dark brown lines (Figure 5).
170 chapter six
Some of the other possibly early burials in B9a:

23: Male, mature adult, inhumation, N-S, head S, skeleton poorly pre-
served in balk. Furniture (Figure 6): two bracelets with overlapping tapered
ends (452T, 996D); a bone pendant decorated with drilled holes and held
by a bronze loop (764T); a plain ring (D); various paste beads colored blue
and white, some brown stone beads, and a bronze coil (1008P); two tanged
daggers (646T, 649T) placed in a jar, 974; a glazed faience cylinder seal of
Mitannian design, the ends of which are beveled, perhaps indicating an orig-
inal holder (information from Edith Porada): a goat or ibex and two stags
move right, but turn their heads left; stylized plants divide them (637T). Also
one plain squat gray IIC spouted vessel with a ridge at the back (932T), and
a gray IIC jar (974D), both at the head, and a gray IIB bowl with two holes
(892T), at the feet.
24: Mature adult, inhumation, flexed on L side, N-S, head S. Furniture
(Figure 6): a bracelet with overlapping ends on R wrist (453T); two plain
rings (1014P); one pin with incised top on L shoulder (472T), another on R
shoulder (479P), and one with a knobbed head on R arm (477P); a needle by
the wrists (D); a white rams head bead with a blue band (1048M; Muscarella
1968, p. 194, fig. 19);6 scores of various types of beads by the neck: 1052a,
e, f, h, i, j, k, paste; 1052b, copper; 1052d, g, Egyptian blue (T). Two vessels
were found slightly below the skeleton: a buff, matt, spouted vessel with a
relatively high foot, a short, squat spout, and a stylized eye or horn motif
in relief at the rear (936T); and a buff, matt, deep, carinated bowl with one
hole below the rim (866T).
26: Child, inhumation, flexed on L side, N-S, head N (Muscarella 1968,
p. 192, fig. 16). Furniture (Figure 7): two anklets on L foot (603P for one), the
other, 536T, is of the same type as bracelet 532, top; two anklets on R foot
(534T, 535T), same types as the preceding; two bracelets with overlapping
ends on R wrist (532P, top and bottom); two on L wrist (531T); a plain ring
with overlapping ends (600T) on L hand; a bronze bead at throat (468T);
two plain pins with blunt tops near throat (478T, 607T); a needle (606D); a
torque of twisted wire and bent-back ends and with a twisted loop attached
(639P); a bronze plaque pierced with a large central and four smaller corner
6 Tests on the bead were made by J.H. Frantz and Suzanne Heim in the Research Labora-
tory of the Metropolitan Museum of Art. The tests show that the bead is not glass, leaving the
following possibilities: a glazed soft-stone, faience, or glazed earthenware. Only the surface
and the inlay appear to be vitreous. Compare A. von Saldern, Other Mesopotamian Glass
Vessels (1500600bc), in Glass and Glassmaking in Ancient Mesopotamia, ed. A. Leo Oppen-
heim (New York, 1970), p. 217.
holes (599P); broken plain rings found by the teeth (D); two gold earrings
consisting of a cluster of hollow balls with a loop: one was found by the
left leg, the other under the skull; associated with the earrings are gold
loops (629T); 76 flattened carnelian beads, 21 similar-shaped copper ones,
plus 150 round paste beads; and one calcite disc (622T), at the back of the
neck. Vessels included a bronze omphalos (542T) by the chin; a broken gray
burnished bowl (881D); a broken, buff, smoothed basket-handled teapot
with mica flecks (792T), and a broken, gray-brown burnished spouted vessel
(922D), same type as 234 in Figure 16, by the feet.
27: Male, mature adult, inhumation, flexed on back, N-S, head S. Right
arm bent back to touch its own shoulder (Figure 8). Furniture (Figure 7):
plain bracelet with overlapping tapered ends (605D) on R wrist; a stone
button with drilled designs (616P) by L foot; assorted beads by throat: 833a,
coarse faience; b, fine faience (glass?); c, paste; d, e, f, carnelian; g, a lotus-bud
shape, fine faience; h, j, glass (P); also, a socketed spear on L leg so that the
shaft must have passed over the body (1045T; compare Dinkha II burials B9a,
g, and B10a, 12, Figures 24, 36). At the feet, a dark gray burnished spouted
vessel (921D, same as 234 in Figure 16), and a gray IIC worm bowl with two
holes (889P).
We now proceed to some of the other Dinkha III burials; these do not
allow themselves to be distinguished as early or late on the basis of stratig-
raphy or artifact comparisons. Space limitations forbid publication of all the
burials, but no important features of the period will be omitted (see Table I).
The burials are presented according to their grid positions:
b9a, 15: Female (?), mature adult, flexed, on back, N-S, head N, in hor-
izontal brick tomb (Figure 9). Furniture: one round bracelet with overlap-
ping ends on R arm (307P), two on L (308T, 354D); two plain rings with
overlapping ends, one on R hand (595T), one on L (601P); two pins with
simple knobbed heads at throat (Figure 52, 400P), one was found stick-
ing up in the fill; another pin with one knob by L arm and another by R
(Figure 52, 385T); a needle over the chest (460P); paste and copper beads
(391T); and five bronze buttons found on the skull probably from a cap or
diadem (617P; B8e, 8, a Period III tomb of a mature adult, also had five
bronze buttons on the head. The buttons as shown in the photograph may
be in their original position; there is no comment in the field notes to the
contrary). At the feet, a highly burnished gray spouted vessel with ridges
uniformly arranged around the body (334M; Muscarella 1968, p. 193, fig. 17,
top), a gray IIB bowl with two holes (358T), and a gray IIB jar with two ridges
at mid-body (404T) sealed with a stone; this vessel probably held some
liquid.
172 chapter six
Figure 6.
Figure 7.
174 chapter six
Figure 8. B9a, burial 27.

b9a, 17: Mature adult, inhumation, flexed on L side, N-S, head to N; R arm
missing, L bent back onto its own shoulder (Figure 10). Furniture (Figure 11):
a flattened bracelet with overlapping ends on L (310T) and R (309P) wrists;
an anklet with overlapping ends on each foot (539T, 540P); a plain toggle pin
at L shoulder (326D), fragments of another in the fill; a ring of twisted wire
with overlapping ends on L hand with cloth impression (466T); a needle in
the fill, with top bent back to form the hole (325D); a plain torque on the
neck (1038D). Touching the forehead was a gray IIB spouted vessel (337T);
by the feet, a broken gray IIC bowl with two holes (893P), and a gray IIC
carinated jar (903D).
b9a, 19: Child, inhumation, flexed on R side, N-S, head S; L arm flexed
across body, R bent back onto its own shoulder. Furniture (Figure 12): a
plain, not quite round bracelet with overlapping tapered ends (541T), on R
wrist; a plain ring, also with overlapping tapered ends (462D), on R hand.
A buff IB tripod bowl, feet of which were broken (982D), resting on a gray
burnished jar (952D), at the forehead; at the feet, a buff smoothed carinated
bowl (871D) and a gray burnished goblet (717T).
176 chapter six

Figure 11.
b9b, 11: Child, flexed on L side, N-S, head N; in brick tomb with projecting
arms (Figure 2; Levine 1971, p. 40, top); the fill in the grave was packed in
very hard. Furniture (Figure 13): a plain flat band bracelet with overlapping
ends (369P); a plain round bracelet with overlapping tapered ends on R
wrist (319D), two on L (320P; like Figure 7, 532); two plain loop earrings (?)
(351D); a bronze spiral object (bead? pendant?) at the neck (618P); two plain
flattened anklets with overlapping ends on L foot (311P), two on R (312D); a
shell bead necklace (299P); and a plain torque (538T). Placed at the feet: a
gray IIB spouted vessel decorated with ridges around the upper body (84T),
an orange IIB carinated jar (83T), and an orange-red IIB bowl (85P).
b9b, 12: Child, inhumation, flexed on the back, N-S, head S (Muscarella
1968, p. 192, fig. 15). Furniture (Figure 13): a plain bracelet with overlapping
ends (350D) on R wrist, the sole jewelry. Covering the head was a gray IIB
tripod worm bowl with mica flecks, and one hole (88M; Muscarella 1968,
p. 193, fig. 17, right), in which were three astragals; by the feet was a gray IIB
goblet, missing the handle (87P), and an orange IIB carinated jar (86T).
b9b, 16: Young adult, inhumation, flexed on R side, NE-SW, head S (Fig-
ure 2). Furniture (Figure 14): a plain round penannular bracelet on R wrist
(300D); a bronze necklace with its hooked ends linked together (therefore
not a true torque), with cloth impressions, at the neck (1037T); a necklace of
paste disc beads (301P); a broken bone awl in the fill (242T); and a tanged
dagger in the fill (241P). At the head was a gray IIB spouted vessel with
a hatched design on the base, exhibiting mica flecks (269T), and a gray
smoothed jar, also exhibiting mica flecks (949D); at the feet was a gray IIC
tripod bowl with two holes (359T).
178 chapter six
Figure 12.
Figure 13.
180 chapter six
Figure 14.
Figure 15. B10b, burial 10.

Figure 16.
b10b, 10: Female, mature adult, flexed tightly on L side, N-S, head N; in
horizontal brick tomb (Figure 15; Muscarella 1968, p. 189, fig. 2). Furniture
(Figure 16): a plain pin at L and R shoulder (138P, 137T); a pin, square in
section, with the top twisted into a loop, at R shoulder (200P); a needle
at the chest (198T); plain loop penannular earrings (148T); a flattened ring
with tapering, overlapping ends on R hand (199P); fifty small round paste
and bronze beads at the neck (896T). Clustered at the feet: a gray IIBC
spouted vessel (234P), a gray burnished carinated jar (939P), and a broken
red-slipped worm bowl with two holes (357P); animal bones were found in
the bowl.
182 chapter six
Figure 17.
b8e, 7: Both the brick tomb and skeleton were disturbed. Furniture
(Figure 17): an orange matt basket-handled teapot with a broken spout
(937T), and a buff IIC worm bowl with two handles and two holes (891D).
Only these two vessels were found.
Dinkha III and Hasanlu V
Not enough material from Hasanlu V has yet been published to permit a
comprehensive comparison of the material from both sites. As more Has-
anlu material becomes available we will no doubt recognize more parallels
and connections than are given here.
The sites are about fifteen miles apart, separated by ridges, but with no
impediments to travelers from one site to the other. That travelers, mer-
chants, and perhaps potters and other craftsmen did indeed travel freely
and often back and forth is documented by the obvious strong ties between
the sites, evidenced by the material culture that was basically the same in
many cases, and very close in others (Muscarella 1968, pp. 189, 194). And not
only does this closeness obtain in the Iron I period, but, as will be seen, it
continued throughout the succeeding Iron II period. Differences did in fact

exist, but collectively they cannot alter the suggestion that there was a basic
identity of culture at Hasanlu and Dinkha.
burials: Both at Hasanlu and at Dinkha the dead were usually buried
in an extramural cemetery in a flexed position with no special orientation.
Similar types of pottery, bronze jewelry, and parts of animals for food were
deposited in the graves (Dyson 1965, p. 196; 1967, p. 2957; Stein 1940, pp. 397
404). At Dinkha, however, some of the burials were placed in mud-brick
tombs, a feature not recorded at Hasanlu, where simple inhumations were
the rule. Also, seven of the Dinkha III burials contained torques; at Hasanlu
only some Period IV graves contained torques; none were found in Period V.
At least one Hasanlu V burial contained a skeleton that held a vessel in
its hand (Stein 1940, p. 402); one burial at Dinkha (Muscarella 1968, p. 192,
fig. 15) held a bowl that was placed on its chest.
pottery: The Hasanlu V wares were characterized by burnished or
smoothed gray and buff surfaces, including red-slipped pottery; but whereas
at Hasanlu (in both Periods IV and V) buff surfaces predominated over gray
by about 60 percent to 40 percent, at Dinkha III gray surfaces predominated
(see above; compare Dinkha II below; Dyson 1965, p. 198; Young 1965, pp. 55,
57; Stein 1940, pp. 401402). Rare examples of patterned burnished vessels
occur at Hasanlu V, but do not occur in the graves of Dinkha III (two such
sherds were found in the fill of the lowest Iron Age trash).
Painted pottery was rare at Hasanlu V but occurs in the form of black or
red-brown bands on a buff ground (Dyson 1964a, pp. 3637, fig. 3:3, 6; Young
1965, pp. 55, 57, 67, fig. 8, 70ff.; these seem to be rare examples of continuity
from the Bronze Age). A remarkable and close parallel to the only painted
vessel found at Dinkha (Figures 3, 5), was excavated by Stein at Hasanlu in a
Period V burial (Stein 1940, p. 401, fig. 110, pls. xxiv, 3, xxxi, 8), neatly adding
to the evidence for strong contacts between the sites.
Bridgeless spouted vessels, pedestal-base goblets, worm bowls, jars, and
carinated bowls are all recorded at Hasanlu (Dyson 1962, p. 5, fig. 4; 1964a,
pp. 3639, fig. 3; 1965, pp. 195196, fig. 17; Young 1965, pp. 57, 7072, 67, fig. 8).
But whereas at Hasanlu V no spouted vessels are reported from burials
they occur only on the moundat Dinkha they were found in burials.
Dyson (1965 p. 196) originally suggested that the presence of spouted vessels
in Hasanlu IV graves, and also at Geoy Tepe, was a late development; the
Dinkha evidence contradicts this suggestion as a general rule.
At least one example of a bowl with an eye/horn motif seen in Fig-
ure 6, and at least one example of a basket-handled teapot, both unpub-
lished (but see Stein 1940, pl. xxiv, 1), occur at Hasanlu V. But bowls with
184 chapter six
vertically pierced handles, jars with one handle, and cups like those found
in Hasanlu V (Dyson 1965, fig. 13; Young 1965, pp. 7273, fig. 11), do not occur
in the Dinkha graves.
jewelry: The types of pins, bracelets, anklets, etc., from Hasanlu have yet
to be published. We can state, however, that torques were not found in the
Hasanlu V graves and that a few toggle pins were found (Dyson 1968a, p. 23).
Two Dinkha graves contained bronze buttons or studs that belonged
originally to a headband, diadem, or cap (Figure 9). At Hasanlu a Period V
burial (VIF, 8) contained a plain bronze band, curved to fit the head and
pierced at both ends, presumably for attachment to another, perishable,
material. Headbands were also reported from Period IV graves at Hasanlu.
The gold earrings from B9a, 26 (Figure 7) are similar to gold earrings
found in Hasanlu IVs Burnt Building II, attached to an ivory statuette frag-
ment (Muscarella 1966, pp. 134135, fig. 36). This earring has the cluster of
hollow gold balls, but placed under a buttonlike form attached to a twisted
gold wire. Another gold earring, consisting of hollow carinated balls in a
pyramid cluster, and attached to a loop, was found at Hasanlu in 1947 (Rad,
Hakemi 1950, fig. 90b). This earring is more elaborate than, but related in
form to, the Dinkha earrings.
weapons: Four of the Dinkha III burials contained weapons, a spear and
four tanged daggers in all. In the same Period V burial at Hasanlu that
contained the painted jar, Stein (1940, p. 402, pl. xxvi, 2) found a bronze
spear; and Dyson (1964a, pp. 3435, fig. 2:1) published a bronze dagger with
a lappet-flanged hilt (57129) that came from a Period V burial.
As stated above, iron was not found in any Period III burial. At Hasanlu
only one iron ring was found in a Period V context (Dyson 1964a, p. 39; 1965,
p. 196; 1967, p. 2957).
Two Dinkha III burials each had among the grave goods three astra-
gals placed in a bowl. Were they from meat, or were they game pieces?
There is certain evidence in Period II at Dinkha that astragal game pieces
were placed in tombs (see below), but it seems to me that in these cases
the astragals were probably simply the remains of meat placed as food in
bowls.
Dinkha III and other Iron I Sites
solduz: Several mounds surveyed in the vicinity of Hasanlu are reported

to have Iron Age gray ware (Dyson 1965, p. 196), but it is not certain if they
belong to both the Iron I and II periods (Young 1967, p. 22, note 70). One
of these sites, the late neolithic or chalcolithic mound of Dalma Tepe, just
south of Hasanlu, had a number of Iron I burials deposited in its upper
level. One grave, Operation IV, 2, contained a gray pedestal-base goblet
(MMA 62.173.9; Young 1962, pp. 707708, fig. 8), but little more can be said at
present than that an Iron I extramural cemetery existed here and that there
may have been a settlement somewhere in the vicinity.
In 1968 at Hajji Firuz, a neolithic mound southeast of Hasanlu, an Iron I
inhumation of an adult was found (unpublished). The arms and legs were
flexed, and the body was oriented NNW-SSE, head NNW. The grave con-
tained all the classic diagnostic pottery of the period: a bridgeless spouted
vessel, a worm bowl, and a goblet (information from Mary M. Voigt). It
will be remembered that none of the Iron I burials at Dinkha (or at Has-
anlu) contained all three diagnostic vessels together (but see Geoy Tepe,
below).
geoy tepe: The B period, in particular the contents of a single tomb,
Tomb K, represents the sole published evidence for an Iron I occupation
here (Burton-Brown 1951, pp. 141ff., figs. 28, 29, 32, 34; Dyson 1965, p. 196,
fig. 2; Young 1965, pp. 7072, fig. 11, p. 78). Recent tests made at the site
indicate that Iron I trash deposits occur; thus evidence of occupation from
that period exists although never extensively excavated (personal commu-
nication from Robert H. Dyson, Jr.). Geoy Tepe B shares with Dinkha III
the bridgeless spouted vessel, the goblet, and the worm bowl, all found
together; in addition, there are toggle pins very close to those at Dinkha
(Figure 11, 326), and tomb architecture, albeit not bricks but stone (compare
Dinkha II).
haftavan: Here settlement on the mound seems to be indicated along
with an extramural cemetery just below the citadel, as at Hasanlu. In the
settlement area were found bridgeless spouted vessels and worm bowls; and
a spouted vessel and a cup were found in an inhumation burial (Burney 1970,
p. 170, figs. 8:1, 7, pl. iii; 1973, pp. 155, 162164; Burney-Lang 1972, fig. 40). No
more data are presently available.
yanik tepe: No settlement occupation was found, but a cemetery at the
foot of the west side of the mound was located. Here eight Iron I burials were
excavated, of which only one has been published, A6 (Burney 1962, pp. 136,
146147, pl. xliv, figs. 2429). The flexed bodies have no particular orienta-
tion and were placed on the left or right side; the graves were sometimes
lined with mud brick, a feature at present recognized in the Iron Age to my
knowledge only at Dinkha. No spouted vessel occurs in A6, but a vessel with
a vertical loop handle, similar to the Iron I goblets, was a clue to the graves
approximate date; this burial may belong to a late stage of Iron I.
186 chapter six
A burial from Trench P, in which were found toggle pins and a painted
jar, along with two hand-made vessels, was dated to the Iron I period, about
1000bc (Burney 1964, p. 60, pl. xv, 1419); this date is not certain, but if
correct, the grave is then surely of the Iron II period. Toggle pins, for example,
occur in the Bronze Age and throughout the Iron Age.
tashtepe: Dyson (1965, p. 196) referred to Iron I pottery from Tashtepe
based on Ghirshmans claim (1954, pp. 6162) that gray wares similar to those
from Giyan were found there on survey. These sherds remain unpublished
and therefore prevent independent acceptance of Ghirshmans statement
(compare Young 1967, p. 22, note 70).
khurvin: This site was plundered by local inhabitants and only a few
graves were excavated by vanden Berghe (1964, pp. 6 ff.). The graves are not of
Iron I date (Dyson 1965, pp. 196, 206) although a particular type of bridgeless
spouted vessel of late type is common; Goff Meade (1968, p. 125, note 50)7
dated the burials to Iron II based on analogies with Sialk B. True Iron I
vessels said to be from Khurvin, but without archaeological contexts, exist
in private collections (vanden Berghe 1959, pp. 123124, pl. 153, and p. 124,
pl. 158 for Chandar; 1964, passim; Ghirshman 1964, figs. 15, 16); others are
from controlled field surveys (Young 1965, fig. 9). The bridgeless spouted
vessel of Iron I type, goblets, bowls on tripodssimilar to worm bowls,
but without the wormare part of the repertory. Also reported are familiar
bronze torques, tanged swords, needles, toggle, incised, and plain pins, and
pins with curled tops (vanden Berghe 1964, pls. iv, xiv, xv, xxiii, xxvi,
xxxiv, xxxix, xli, xliii). While we cannot control the information enough
to actually know if this metal material is Iron I rather than later, given the
material itself and the pottery configuration, it is quite possible that part of
it, at least, is early (Moorey 1971, p. 25; for pins and needles, pp. 172215).
Vanden Berghe also relates (1964, p. 3) that the burials were all inhuma-
tions without any particular orientation; presumably he is talking about the
burials he excavated, but he implies that he is also discussing those burials
dug by the local inhabitants.
marlik: Without doubt some of the Marlik tombs belong to the second
millennium bc while others must be later (Muscarella 1972, pp. 4243).
Bridgeless spouted vessels in metal and pottery, as well as toggle pins and
tanged swords, are attested there (Negahban 1964, figs. 25, 29, 41, 108, 121,
7 Claire Goff (Meade) considers the unbridged spout with curled ornament to be late
Iron II, eighth century bc (1968, pp. 115, note 17, p. 121), while Dyson considers it to be Iron III,
eighth century bc (1965, p. 206, fig. 11), and Young (1965, p. 73, fig. 11) lists it as Iron I. The
evidence from Sialk B suggests that it was in use in the eighth century.
135). Two published gold earrings are not dissimilar to the clustered hollow
balls on examples from Hasanlu and Dinkha.
kizilvank: Bichrome vessels closely paralleling the Hasanlu V and
Dinkha III painted vessels discussed above were excavated here (Schaeffer
1948, p. 500, fig. 270; Muscarella 1968, p. 194). Moreover, aside from the spe-
cific paint parallels, two of the bridgeless spouted vessels have a short spout,
one has a rather narrow mouth, and both have a straight interior wall on
the vertical spout, all features in evidence on the vessel from B9a, 24 (Fig-
ure 6). A goblet from the site is similar to those of Iron I type, but it has a
flat base. Schaeffer (1948, p. 500) dated these vessels on typological grounds
to between 1350 and 1200bc. Monochrome red and gray pottery of the Early
Iron Age, bronze daggers similar to those from Dinkha III, and a flanged dag-
ger of Iron I type are reported from the site (Burney-Lang 1972, p. 169, fig. 43a,
b; compare Dyson 1964a, figs. 1:5, 2:1, and p. 34).
sialk a: Moving to the south, to central western Iran, we see that the Iron
Age culture extended as far southeast as Sialk and as far south as northern
Luristan (Goff Meade 1968, pp. 127132; compare Dyson 1968a, p. 25, for a
similar situation existing in the Late Bronze Age).
The necropolis of Sialk VI, Necropole A, provides the relevant informa-
tion (Young 1965, pp. 6162, 73, fig. 11). Here only an extramural cemetery
is available for study. Some Iron I vessels, however, do come from limited
excavation on the South Hill, where we are also told related architecture was
cleared (Ghirshman 1939, p. 11).
The skeletons were flexed in single burials, with no particular orientation,
except that most of the heads pointed north, as at Dinkha. The pottery
covers the range of familiar Iron I wares and shapes (Young 1965, pp. 6162;
Dyson 1965, p. 195). The clothed dead wore bracelets, pins, rings, and at least
one needle was found. One tomb contained gold; another, probably late (see
also Moorey 1971, p. 316), an iron tanged dagger and an iron point, along with
bronze weapons (Ghirshman 1939, pl. xxxix). Note that Young (1965, p. 62)
suggests that Necropole A lasted a long time.
giyan: Young (1965, pp. 62ff.) has reorganized the subdivision for Giyan I,
a system accepted by Dyson (1965, p. 195, note 5). At Giyan we have basically
a cemetery with no definite related settlementexcept it is possible that
Construction A may be contemporary to some of the burials, but this is by
no means certain (Young 1965, p. 66). Graves of Giyan I4I2 are the ones of
concern to us.
These graves are simple flexed inhumations with no apparent orienta-
tion. Except for one bridgeless spouted vessel of a type also found at Sialk
B and at Khurvin, and which may be later than Iron I, the shape is not
188 chapter six
represented at Giyan (see note 7). The pedestal-base goblet is fairly com-
mon, however, especially in I4 and I3. The dead were buried with bronze pins,
needles, bracelets and anklets, and occasionally with a tanged dagger, in one
case (late?) iron. One skeleton wore a headband of bronze loops, and a sin-
gle cylinder seal was found (Contenau, Ghirshman 1935, pp. 23, 26, pls. 14,
18). Another seal, of Mitannian type, was found low in Construction A and
could have come from a tomb.
godin: Three isolated burials containing Iron I material, but with no rela-
tionship to any settlement on the adjacent Godin mound, were discovered
in a Bronze Age cemetery (Young 1969, p. 19, figs. 24, 25). They are all simple
flexed inhumations, oriented E-W, on their R or L sides, facing N or S. Each
grave contained a typical Iron I goblet. One grave contained a ring and two
pins; another a bracelet and a bronze cup; the third an arrow and a sword
with an open crescent handle. Two skeletons held vessels in their hands
(compare p. 48 above).
Interestingly, each of the goblets is slightly different in base type and
outline, which does not necessarily signify that they were deposited over
a long period of time. It should also be noted that several toggle pins with
decorated tops from a Godin III, Bronze Age, burial (Young 1969, fig. 30) are
quite similar to an early example from Dinkha (Figure 3, 473).
tepe guran: In an occupational context of Level VII, the latest settlement
at Guran, a bridgeless spouted vessel with a handle (unlike Dinkha) was
excavated (Thrane 1964, pp. 122, 123124, figs. 23, 24; 1965, pp. 158159, note 6).
Cut into this level, and therefore later, was grave 4, which contained a bronze
spouted vessel of a type similar to those from Hasanlu IV and Sialk B (Thrane
1964, p. 129, figs. 30, 31; 1965, pp. 158159, note 6; Moorey 1971, pp. 276280).
Thrane, nevertheless, dates Level VII to the Sialk B period, that is, to the early
first millennium bc (also Thrane 1970, p. 31, 850750bc; Moorey 1971, p. 21).8
It would seem that the stratigraphically later grave 4 is Iron II in date, and
that Level VII may belong to the Iron I period.
A word should be said about the rams head bead from B9a, 24 (Figure 6,
note 5; Muscarella 1968, p. 194, fig. 19). Similar frit and glass beads were
found at Nuzi (Starr 1939, pl. 120), Alalakh (Woolley 1955, pl. lxviii), and
al-Rimah (Carter 1965, p. 51), all approximately mid-second millennium
8 Moorey inadvertently placed the bridgeless spouted vessel from Thrane 1964, fig. 24, in
Grave 4 (Moorey 1971, p. 21); in fact, it was found in situ in Level VII (Thrane 1964, pp. 122131,
figs. 23, 25; 1965, p. 159, note 6; 1970, p. 31, fig. at top). As stated in the text, Grave 4 is related
to Sialk B material and is later than Level VII, making the latter possibly pre-nintheighth
century bc, probably Iron I as argued here.
bc in date. Perhaps we may consider the Dinkha bead an import from

Mesopotamia.
It is also of some interest to note here that brick-lined burials have been
excavated in southeastern Iran at Shahr-i-Sokhta (R. Biscione et al., Iran, XI,
1973, p. 204, pl. xib), dating to the mid-third millennium bc.
From the foregoing summary we see clearly that Dinkha III has its clos-
est ties with Hasanlu V. The pottery and pins from Tomb K at Geoy Tepe,
albeit evidence from one tomb, suggest close ties between Hasanlu and
Dinkha and the western shores of Lake Rezaiyeh.9 The same characteristic
vessels occur still further north at Haftavan, demonstrating that the culture
extended to the northern part of the lake (Young 1967, p. 22, for informa-
tion that no Iron I wares have been found north of Lake Rezaiyeh).10 That
it also existed, or at least was known, on the eastern shore is documented
by the finds from Yanik Tepe. However, the little information published
to date from this site makes it impossible to evaluate how strong the ties
were between the Yanik area and the southern Urmia basin (compare Bur-
ney, Lang 1972, p. 117). One must keep in mind the perhaps significant fact
that Yanik is the only Iron I site other than Dinkha where brick tombs
exist.
Marlik, further east, seems to be in part a contemporary culture with
some ties to the west. But without doubt, Marlik remains a unique and indi-
vidual center (Moorey 1971, pp. 2324; Dyson 1965, p. 211). Future publication
of the tombs and contents should give us more information about the begin-
ning and terminal dates of the tombs.
An Iron I settlement with fairly close ties to the Urmia basin must have
existed close to the Khurvin-Chandar cemeteries. Of interest, aside from
the pottery and metal ties with the northwest, is the occurrence of bronze
torques. As stated, we do not know if they are Iron I or II, but the former is
not excluded (Moorey 1971, p. 229). The occurrence of torques in burials is a
continuation of an earlier widespread custom (Schaeffer 1948, p. 111, figs. 53,
56, 58, 59, 134, 194, 544, pls. xv, xvi; 1949, pp. 49120 with reference to the
Near East and Europe; Moorey 1971, pp. 229230), one that continued into
the early first millennium at Dinkha II and Hasanlu IV and still later into
the Achaemenid period.11 Besides its use at Dinkha III and II, Khurvin, and
9 The only anomaly at Geoy Tepe is the use of a multiple burial, whether or not we
interpret it as a one-time deposition or a result of continued use.

10 Note that a single nipple-base goblet of Iron I type is said to have been found at
Toprakkale; H.Th. Bossert, Altanatolien (Berlin, 1942) fig. 1201.

11 See note 5. Burton-Brown 1951, p. 6, note 5, Schaeffer 1948, p. 544, note 1, and Schaeffer
190 chapter six
Hasanlu IV, the torque was used in Luristan (Godard 1931, pl. xxvi, 78, 80),
at Sialk B, and in the Caucasus and Talish regions (Schaeffer 1948, figs. 254,
298, 301; Hancar 1934, p. 97; Godard 1931, p. 64, fig. 34; Barnett 1967, pp. 177,
174, fig. 27:3; Herzfeld 1941, p. 146, pl. xxx, says some were found at Giyan).
We do not really know if the torque was first used in the Talish area, or in the
Urmia basin areawhence it could then have moved north and southor
whether the Khurvin examples are contemporary to those at Dinkha, having
been a basic element in the Iron I culture from the first years of settlement
in Iran.
The painted pottery from Kizilvank is difficult to evaluate. One asks:
does this site represent the first stage of the new Iron Age in its incipient
phase, thus affording us a clue about the area of origin (see Burney, Lang
1972, p. 116), or was it a backwater, being later than, or even contemporary
with, the Iron I culture to the south? It is preferable to leave the questions
unanswered at this stage.
The people in the central plateau, at Sialk, Giyan, and Godin, had similar
burial practices and included artifacts in their graves similar to those found
further north. Tombs at Sialk and Giyan contained tanged daggers of the
same type found at Dinkha. But they also occur later at Sialk B (Ghirshman
1939, pls. l, lvii, lxviii; Moorey 1971, pp. 6668 for a late dating for some
examples). In the north this tanged dagger is clearly earlier than examples
with cast hilts.
Finally, we have the three graves from Godin. One wonders if they are
in fact isolated and were deposited by a people on the move (as Young
1969, p. 19), or whether there are other burials at Godin still unexcavated
that might perhaps indicate a nearby settlement, or more intense use of the
area. Without any more information at hand the graves offer us merely a
tantalizing glimpse, rather than a substantial view, of the Iron I period at
Godin.
1949, p. 109, refer to heavy bronze torques from Geoy Tepe and Iranian Azerbaijan, based
on a report from C.C. Lehmann-Haupt. These objects cannot be the same objects we call
torques that were found at Dinkha and Hasanlu. Ghirshman 1964, p. 113, fig. 148, following
Godard, refers to a gold fragment allegedly from Ziwiye as a torque, but this is not certain. For
Achaemenian torques see J. de Morgan, Dcouverte dune Spulture Achmnide Suse,
MDP VIII (Paris, 1905) pp. 4344, pl. iv; E.L.B. Terrace, Sumptuary Arts of Ancient Persia,
Boston Museum of Fine Arts Bulletin 13 (1965) p. 27, with references; see also Schmidt 1970,
pp. 111116, and my comments in a review of Schmidt in AJA 75 (1971) p. 444. Note that a torque
with twisted ends, similar to Figure 32, B9a, 14, 1040, seems to be worn by a youth on a relief
from Marash: E. Akurgal, The Art of Greece (New York, 1968) pl. 29.
It has been stated many times that the Iron I culture represents a new
phenomenon in western Iran, a major break with the past and a new age.
The information available from the Dinkha excavations reinforces this con-
clusion both from stratigraphical and cultural evidence. There is a definite
break, a hiatus, after the termination of the last Bronze Age settlement. A
build-up of debris and erosion material covered this destroyed settlement,
creating a hard-packed, easily distinguishable stratum. Ash layers, debris,
and erosion material, containing Iron I sherds, coming from the southeast,
and thinning at the north, were laid down. It was into these layers that the
first Iron Age graves were deposited (Figures 18, 19, 20). Perhaps these Iron
Age layers came from the earliest Iron Age occupancy of the mound, from
a time before the primary use of the area as a cemetery (it will be recalled
that one of the earliest burials came from a test trench in grid L to the west,
TT VII, 2). In any event, trash and ashes continued to be deposited during
the Iron I and II periods.
Culturally the break is equally clear and dramatic, notably in the pottery
and in the burial customs, where single inhumations in an extramural ceme-
tery replace intramural multiple burials. That this new culture represents a
shifting of population, to quote Dyson, and that it represents at the same
time a cultural uniformity, pointing to a common origin for the Iron I cul-
tures, to quote Young, is beyond dispute. Indeed, all the bricks are not yet
available for archaeologists to build a fine structure of full understanding
about the nature of the historical events leading to the change. But further
excavations, conducted scientifically, will continue to supply the necessary
information and slowly put into focus the picture we all seek.12
An item of some importance is the fact that at Dinkha we have been able
to isolate a few burials and their contents that are of the early Iron I period.
At Hasanlu, aside from the polychrome vessel excavated by Stein, we have
no recognizable early material. Thus, although it would be rash at present to
conclude that the Iron I period began earlier at Dinkha (we still do not know
what is in the unexcavated ground at Hasanlu and at other unexcavated
local mounds) we can at least illustrate the earliest excavated Iron I material
there.
12 Excavations by the Archaeological Service of Iran at Gheytareh, north of Teheran, have
yielded a cemetery of Iron Age date: K. Fard, Fouilles dans les Tombes ancien de Gheytareh,
Bastan Chenassi va Honar-e Iran 2 (1969) pp. 2630. A pedestal-base goblet of Iron I type,
said to come from Kal Dasht, near Saveh, southwest of Teheran, is in the Teheran Museum,
no. 872.
192 chapter six
Figure 18. East section, Bgb; B10b in background.
Figure 19. East section, B10a. The Urartian site of

Qalatgah is faintly visible at the base of the distant hills.
Figure 20.
194 chapter six
As sharp as the break was with the past, it seems almost certain that the
Iron I people had some knowledge of the earlier cultures, probably from
scattered pockets of survivors in the penetrated areas. The use of multiple
burials at Geoy Tepe, and the use of toggle pins and painted wares (perhaps
also of gray and red-slipped wares?), reflects a continuity with the past
within Iran, even if not of major proportions.
chronology: A C14 sample from the terminal Bronze Age deposit gave a
reading of 143552 bc (P-1231, half-life of 5730 years; Dyson 1968a, p. 22). This
gives us a rough terminus for the end of this settlement and an ante quem
non date for the following Iron Age.
Three C14 charcoal samples exist for Dinkha III; each came from separate
pits overlying the Bronze Age deposits, from the Iron Age fill. One gives
a reading of 114637bc (P-1475); the second, stratigraphically earlier than
the first, gives a reading of 130257bc (P-1474); and the third from a pit
resting directly on the Bronze Age level, a reading of 1243 37bc (P-1449).
These readings suggest a general date for the end of the Bronze Age in the
fifteenth century bc, and a range of about the late fourteenth to the late
twelfth century for at least part of the overlying Iron Age I burials. They also
suggest that the gap between the end of the Bronze Age and the beginning
of the Iron Age may have been about one hundred years. To be sure, the
pit samples do not necessarily date the earliest nor the latest Iron I burials.
(Note that if the recently published MASCA correction dates for C14 readings
prove to be stable, it will be necessary to push all the dates further back in
time [Ralph et al. 1973, p. 11 and passim]. Thus, the end of the Bronze Age
will have occurred about 1600 bc, and the beginning of the Iron Age about
1500bc. And this correction factor would then change all the dates presented
here by 100 or more years.)13
Dyson (1968a, p. 31) suggested a working date of 1350 50 bc for the
beginning of the [Iron I] period, close to the date of 1300/1250 of Young
(1965, p. 83; 1967, p. 12). This date, about 1350 bc, is also proposed by Burney
(Burney-Lang 1971, pp. 106, 113, 115117; also Muscarella 1968, p. 196). Thus
the tombs that I suggest are the earliest at Dinkha, B9a, 25 and VII, 2,
would presumably have been deposited in the late fourteenth century bc,
13 I use the standard C14 dates in the present report and will continue to use them
until more information and discussion on the MASCA corrections are available. If these
corrections are eventually proven correct, the dating of the many Iron II objects from the
destruction level at Hasanlu IVthe ivories, bronzes, the gold and silver bowls, not to
mention the pottery and architectureshifts dramatically from a late ninth- to a late tenth-
or early ninth-century date, with important implications also for the dating of much material
not from good archaeological contexts.
at least close to 1300bc. Later than these would be burials B9a, 23, and
24, followed by B9a, 26, and 27, and perhaps we could accept a general
thirteenth-century date for these in the order given. The other burials do not
allow themselves to be defined more precisely and presumably span some
centuries, if we can accept the fact that a certain conservatism obtained in
the middle and later stages of the Iron I period.
The terminal date for the Iron I period can be determined at present only
by reference to the large amount of data from Hasanlu. The evidence there
suggests that in the eleventh century bc (or earlier, given the MASCA correc-
tions) major developments occurred on the Hasanlu mound: the building
of fortification walls and large structures, and an expansion of new pottery
forms (Dyson 1965, pp. 197199, 211; Dyson 1968a, pp. 3132; Young 1965, p. 82;
1967, p. 24).
Dinkha II: Architecture
Evidence for Iron II architecture was found in several areas of the mound: in
the main cemetery area and in squares G9ac.
The architecture in the main cemetery area consisted of three kilns and
fragments of walls and rooms. The walls were much destroyed by stone
gathering and burial activities and therefore no complete structure was
preserved.
In square B9b one section of a wall was found under 17 (Dinkha II period,
Figure 2), a fact that established the existence and abandonment of some
structure here before the burial was deposited.
In square B9a in stratum 2, near the surface, a kiln was excavated (see
below). It had evidently been cut into a stratum in which there were at
least two structures; one, at the northeastern corner of the trench, enclosed
by Walls A and B, the other at the northwestern part called Area 1. In the
southwest area of the trench, and a level or two below these structures,
was a single wall with a threshold preserved, called Wall C. It had been
cut into by 4, which in section was seen to be partly under some stones
from Wall B. Urn 5 was under the room area formed by Walls A and B,
and urn 6 was under the kiln. The sequence here would appear to be:
Wall C, followed by 4, and 5; then Walls A and B, and Area 1, and finally the
kiln. Burials were found in all the strata of B9a, but it is not clear just what
the relationship of the structures was to the use of the area as a cemetery,
chronologically and culturally. It may be that structures not considered
convenient for inclusion in a settlement area were built in the cemetery
196 chapter six
Figure 21. Kiln, B9a.
area. Fred Matson has suggested that the kilns were built here because they
would have represented a fire hazard if they had been near a residential
section (compare Stein 1940, p. 394, for a kiln near the Hasanlu cemetery;
compare also the biblical Potters Field, and the Athenian Kerameikos).
The kiln of stratum 2 was roughly oval in plan with a hard earth floor, over
which was an ash layer, and a wall of vertical bricks set on edge (Figure 21);
its entrance faced south. Within the chamber, originally domed, was a firing
unit touching the east wall: a N-S wall of bricks (about .40 .35 .12 m.),
made up of two rows of four bricks each laid flat and with an upright at each
end, that abutted a small chamber of two upright bricks supporting a brick
and a half, which forms the roof. Nothing was found inside the kiln.
Figure 22.
Square B10a had two architectural features, a kiln and the remains of an
interesting building called Structure A (Figure 22). The axis of Structure A
was roughly N-S. Its S wall, A, extant length 5.5m., consisted of large outer
stones filled with smaller ones; the eastern most preserved stone was a
threshold. To the south of Wall A was a pebble pavement, partly preserved,
which was perhaps an outside area. On this pavement were found two buff
IA pots (on the plan 16: 109D, 18a: 104D). Abutting Wall A at the north was
a rubble packing in an L-shape that was obviously a stairway support. One
meter further to the north of the stairway was a series of stones set on edge
bordering a stone paved area, .35 to .50m. in width, that in turn bordered
another, wider, paved area of larger stones 1.30m. in width. Another wall, B,
bordered both the paved unit and the stairway at the east and joined Wall A;
thus it separated the stairway-paved area from a room to the east that made
use of the threshold of Wall A. The paved area seems to be a unit consisting
of a jube (a water channel) set next to a narrow pavement, placed within
a roomor courtthat also contained a stairway to a second story. The
juxtaposition of jubes and pavements exists at Hasanlu IV in the area just to
the west of the fortification walls, in the northwest quadrant of the citadel
area.
198 chapter six
Figure 23. Kiln, B10a.
This structure was built and abandoned before the kiln was constructed.
It seems also that some urn burials were deposited in the area after the
abandonment. Stone tomb 15 is partly under the stone pavement at the
south of the structure, but we cannot be certain that the tomb was earlier: it
may have been later and the burial pit could have undercut the remains of
the pavement.
The kiln was dug into the fill of level 2, just below the top soil (Figure 23).
In plan it was a rough oval built of clay and apparently originally domed.
Its entrance on one of the long sides faced southwest. Within the chamber
were two units, a lower chamber for firing, and an upper one for the pots,
both now collapsed. In the center of the lower chamber was a pillar of three
bricks with a single brick on end touching them; this helped to support the
upper chamber. The floor of this upper chamber consisted of large bricks or
slabs, one of which was found on edge, having slipped. Holes in the floor of
the upper chamber were made to carry the heat to the pots.
Within the chambers was found a broken buff IA jar with three nipples
on each side (107D), and inside the jar was a fragment of a plain bronze ring.
On top of the collapse was a broken, buff spouted vessel (13T).
Figure 24. G9a, b, c; structure. Burials in G10c are apparently Islamic.
In square B7f, a small L-shaped trench, a third kiln was partly excavated.
It too was very close to the surface and was exactly like the kiln in B9a, with
a brick wall and similar firing chamber; it opened to the north, unlike the
other two kilns, which opened to the south.
In squares G9a, b, and c, we excavated the poorly preserved remains of a
large building; recent stone gathering and burial activity had badly denuded
this part of the mound. The building as preserved consists of two rectangular
rooms bordered at either end by smaller squarish rooms (Figure 24). The
walls, 1.15m. wide, are made of large stones on the outside faces with smaller
stones used as filler, similar to the construction of Structure A in B10a. The
brick superstructure was no longer extant but was made of sun-dried clay
bricks, to judge by the wash adjoining the walls. The two excavated squarish
200 chapter six
rooms are the northern limits of the building and they are of uneven size;
the easternmost one is about 3.60 3.75m. (N-S E-W); the western is about
3.603.20m. There is definite evidence for the existence of a third room to
the west, but very little of it has been excavated.
To the south a rectangular room about 7.40 3.70 m. was cleared; this is
the eastern limit of the building. To its south is a partially cleared area that
probably represents a squarish room, balancing the one to the north. The
total excavated length of the eastern facade is about eighteen m. To the
west of the rectangular room is a partially cleared room that could be either
another rectangular room, approximate in size to its neighbor, or a larger
central room or hall. In the latter case we would expect a balancing rectan-
gular room to the west, in the former case we would expect another rectan-
gular room further west. Thus one could conceive a plan that included three
rectangular rooms, bordered at north and south by smaller rooms, or a cen-
tral hall bordered east and west by rectangular rooms, all bordered by side
rooms. Unfortunately, too little was excavated to carry speculation further.
Exterior and interior doors are no longer extant but surely they must have
existed. Floors were hard-packed earth, and no artifacts other than Iron II
sherds were recovered; there was no evidence of burning.
Of special interest are the two flat stones, about 50 30 cm., preserved in
situ set into the floor of the easternmost large room. The northern stone is
about 75cm. from the northern wall, and about 13 cm. from the eastern wall;
the southern one is about 1.05m. from the southern wall and about 13 cm.
from the eastern wall. These stones clearly appear to be bases for now lost
wooden posts. How many other bases originally existed in between the two
extant ones is not certain, but there could not have been more than four
or five stones as a total number. Was the whole room filled with columns
at one time? It would seem from the narrow width that the answer is no,
although this idea cannot be categorically ruled out. However, it seems
easier to visualize a room with a series of posts set around the perimeter,
posts that may have held a balcony. In this respect one may make a formal
comparison to the posts in the Burnt Buildings at Hasanlu, there set flush
against the walls (Young 1966, figs. 1, 2), but nevertheless probably serving
the same function. Perhaps we may call the Dinkha building a manor, in
the same sense that Claire Goff called the building excavated at Baba Jan
in Luristan a manor. This building, slightly later in date than ours, had a
columned rectangular room about twice the width as the room at Dinkha
(Goff Meade 1968, pp. 112115, figs. 4, 5; 1969, pp. 117122, figs. 24). The Baba
Jan manors rectangular rooms were also flanked by smaller side rooms, and
in plan is not altogether dissimilar to the Dinkha manor.
We may conclude that the Dinkha manor was more than eighteen meters
north to south, and more than twelve meters east to west; that it contained
at least nine rooms, that the walls were thick enough to hold a second story,
and that one of the long rooms had posts or columns. In short, there is
evidence for a major building in the Iron II period at Dinkha.
Burials
Sixty-eight burials of the Dinkha II period were excavated, of which nineteen

were infant urn burials and will be discussed separately. The burials came
mainly from the trenches cut in grid B, but a few were found in G10c and
TT III to the south; none were found in B10B. Note that four burials exist that,
because of inadequate evidence, could not be attributed to either Dinkha III
or II; they are listed separately in Table III.
Twelve of the burials were inhumations, thirty-one were brick tombs, all
of the three-sided type, and six (not seven as in Muscarella 1968, p. 189)
were stone tombs. Continuing the earlier practice, a N-S orientation was
preferred, bodies were placed on the back or sides, arms and legs were
usually flexed. In two examples the skeletons had one arm flexed while the
other was bent back to touch its own shoulder, a practice more common in
the earlier period (Table II).
In the brick tombs the opening was generally to the W, and the body
usually faced the wall, less so the opening or sky (Muscarella 1968, p. 190,
figs. 5, 6; Figures 2, 29, 34, 38, 40). Four of the stone tombs opened to the E,
one (B8a, 5) to the W (i.e., that is where the closing slab was placed); one
(B8a, 1) had a large slab at the N and S. These tombs were rectangular in
plan and constructed of irregular stones (Muscarella 1968, p. 189, figs. 13, 14;
Figures 33, 34, 41, 42, 46). It seems that three walls, and a roof, composed
of large stones, were constructed in place before the burial was sealed by
a large slab with filler stones, thus creating a completely sealed chamber.
Five of the tombs had a stone floor, the other (B10a, 6) a smoothed, hard-
pack floor. Most of the bones in these tombs had disintegrated, leaving only
a few fragments, or nothing at all. Presumably this destruction was caused
by the collection of water in the chamber, water that drained slowly, and
that occasionally froze and then thawed. In the open burials drainage was
faster and the skeletons were not damaged.
Single burials were the rule but four burials contained two skeletons
each. One of these was a mother and infant (B8d, 1), another apparently
a mother and child (B8e, 5II); technically, these could be classified as
202 chapter six
multiple burials. The two other burials were in stone tombs and contained
adults (B8a, 1, B8e, 5I).
As in the earlier period, men, women, and children were buried in the
same cemetery area, with no apparent difference in funeral rites or treat-
ment of the corpse recognized with regard to age and sexexcept that
infants were sometimes buried in urns.
In two burials of old adults arthritic lipping of the vertebrae was noted
(B10b, 7, 8), and in one burial, that of a child, a partially healed hole in the
skull was detected (B10b, 3). One burial consisted of disarrayed bones and
seems to represent a secondary burial (B10a, 13).
Burials were recognized as Dinkha IIIron II in date sometimes by depth,
more often by the nature of the contents. This often consisted of a bridged
spouted vessel (both with and without handles, and with a beard project-
ing from the base of the spout), or a hydria (a medium-sized storage or water
vessel with three handles). In addition to these classic shapes, the various
jars, cups, carinated bowls, deep bowls with animal-head handles, and many
metal objects, jewelry and weapons, many made of iron, and all well known
to us from Hasanlu IV, made attribution fairly easy.
Thirteen of the burials did not have a spouted vessel, but in about eight
of these attribution to Dinkha II could be made on the basis of other shapes.
As was the case with Dinkha III burials, both complete and damaged pottery
were considered as possessions adequate for the dead.
Gray and buff pottery continued to be used side by side. In this period,
however, buff pottery predominated. The total number of vessels from
Period II was two hundred and fifty-two: two hundred and twenty-nine from
the burials, nineteen urns, two from the kiln, and two from Structure A. Of
these, sixty-seven are gray and one hundred and eighty-five are buff (nine-
teen of these are the urns): the percentage of gray to buff is therefore about
27 to 73.
Among the gray pottery, burnished and smoothed surfaces are even,
twenty-eight recorded for each, four are matt, and six were not recorded by
surface treatment. Among the buff pottery fifty-five are smoothed, twenty-
four burnished, sixty-nine (counting the urns) are matt, and ten are red-
slipped; the rest were not recorded by surface treatment. (Thus, as in
Period II, gray vessels were more likely to have been burnished than buff
vessels.) In color, sixty-seven vessels are orange, six are red-orange, eleven
are red; the rest were simply listed as buff.
In both the gray and buff pottery, common-ware paste, with few or no
inclusions visible, predominated about two to one over medium-sized grit.
Only two vessels were recorded as having mica flecks (Mica flecks exist not
only in Dinkha IV and III, but also on the Iron III pottery from nearby Agrab
Tepe, Muscarella 1973, p. 65).
The number of vessels associated with a burial varied from none (usually
incompletely excavated burials) to twenty-six, the majority having four or
five (Table II), and there appeared to be no special relationship between
type of tomb, and age and sex, to numbers of vessels or grave goods. The
only notable exception was that most of the stone tombs, but not all, were
among the richest of the burials.
Thirty-seven burials contained some form of jewelry; torques were found
in four burials. Seven burials contained weapons and only one burial con-
tained horse bits. Jewelry and weapons were made from both iron and
bronze, but the latter clearly predominates. A count14 of the available inven-
tory yields the fact that there are about one hundred and seventy-two bronze
pieces of jewelry and eighty-one of iron, and among the weapons there are
sixteen made of iron and three of bronze. Sheep/goat bones were commonly
found in the burials, and it is possible that liquids were placed in some of the
closed vessels.
We have seen that it was possible to isolate a few Dinkha III burials as hav-
ing been deposited at an earlier stage than other burials of the same period.
In a few cases this differentiation was also noted among the later Dinkha II
burials. However, in these examples the distinction was suspected primar-
ily on the basis of relative depth and does not seem to have independent
support on typological grounds. (Unidentified so far from both Hasanlu and
Dinkha is a transition grave from Iron I to II.) These possible early graves
include B9a, 9; B9b, 19; B10a, 16; and B10b, 11:
b9a, 9: Male, mature adult, flexed on back, E-W, head W; arms at sides
touching pelvis; in brick tomb (Figure 25). Furniture (Figure 26): a plain
round iron penannular bracelet, broken (415P), and a bronze corrugated
band type with overlapping ends (820T) on R wrist; an iron and a bronze
penannular ring, both plain (413P), and two plain iron ones, broken (419P),
all on R hand; a plain corroded bronze pin (383T); also a necklace of car-
nelian, paste, and Egyptian blue beads at the neck (389P). A bronze spear
with short, ovate blade was placed point up along the left side of the head so
that the shaft crossed over the body (221P); and an antler ax with remains of
the wood shaft along with a bronze and iron stud in situ (1042T) was placed
14 It was not possible to give the absolute number of metal objects; some had disintegrated
and were not given catalog numbers, and in some cases rings in the inventory were given
one number, while I counted them as two objects.
204 chapter six

Figure 26.
Figure 27.
206 chapter six
Figure 28.
next to the spear. A broken dark gray IIC spouted vessel, with a horned ani-
mal in relief on both sides (335T), resting on a stone, and a broken carinated
orange IIA jar (904D), were found on the R side; a sharply carinated orange
burnished bowl (870T) was at the head, and two jars one, buff IA (173D), the
other orange IIB (252P), one with a sherd over its mouth, were placed at the
R shoulder.
b9b, 19: Adult, flexed on L side, N-S, head N; in brick tomb. Furniture
(Figure 27): two bronze pins with decorated grooved tops at the shoulder
area (375P, 382P) a third on the chest (374T); two bronze earrings consisting
of a large plain loop with connected hooked ends attached to a smaller loop
(1009M, 1010T), a plain bronze ring with overlapping ends (352D), in the fill; a
group of beads (394T), and a plain bronze torque with bent ends, at the neck
(1039P). At the feet was an orange IIB spouted vessel with an animal-head
handle (238M, Muscarella 1968, p. 190, fig. 9), and at the back was a red-
orange IIB bowl with flaring sides and two holes set within grooves (236T),
and an orange IIB jar (422T).
b10a, 16: Infant, flexed on L side, most of the bones missing; N-S, head N;
in brick tomb partly destroyed by B10a, 6 (see below). Furniture (Figure 28):
two round, coiled, and twisted bronze bracelets, very corroded (181T, 203T);
a simple bronze torque on the neck (187T), along with some beads (439P).
At the feet were a gray IIB basket-handled teapot with mica flecks, in fact,
probably a milk bottle (224M, Muscarella 1968, p. 190, fig. 8), a gray IIB tripod
bowl, with one foot missing (230T), and a small gray IIB jar, also with mica
flecks (255D). (Note, in Muscarella 1968, p. 191, fig. 12, the caption should read
that the vessels came from B10b, 16, not B10a.)
Figure 29. B10b, burial II.
b10b, 11: Female, mature adult, extended on the back with the arms
flexed across the chest; N-S, head N; in brick tomb (Figure 29). Furniture:
a plain, blunttopped, bronze pin at L shoulder (142P) and R (155T) (compare
Figure 7, B9a, 26, 607, Dinkha III); a bronze pin with a hooked end at the R
(154P) and a plain bronze pin (619T: like Figure 16, B10b, 10, 137, Dinkha III,
and Figure 45, B8a, 1, 709, Dinkha II) over R shoulder; and similar pins in
iron, one at R shoulder (444T) and one at L (445P), for a total of six pins. A
flat iron ring with three grooves, on L hand (147P), an iron archers ring on
the R (427P; compare Figure 35, B10a, 6, 195); a necklace of stone (carnelian,
jasper) round beads, others round frit, paste, glass, and copper (827T); one
is a lentoid antimony bead (437P). In the northeast corner, an orange IIB jar
(257P), by the hip a gray IIB one (179D).
At first I was inclined to place this burial in the Dinkha III period, mainly
because there are no other extended burials in the II period, and because
the pottery was not distinctive. But Stein (1940, p. 400) excavated at Hasanlu
an extended burial of Period IV date. Moreover, iron does not exist in any
bona fide Dinkha III burial, and archers rings (iron) occur only in bona
fide Dinkha II burials, and in Hasanlu IV. I therefore believe that this burial
belongs to Period II.
208 chapter six
Figure 30. B10b, burial I.
Finally, there were two burials that were found high in the fill and may
be considered to be later than most of the others (B10a, 1, B10b, 2). One is
published herewith:
b10b, 1: Young adult, inhumation, flexed on R side; E-W, head E; R arm
extended, L arm touches knees (Figure 30). Furniture: a gray burnished
spouted vessel by chest (845T); another spouted vessel, gray IIB, with the
spout broken, and with three nipples in an inverse triangle on each side
and two vertically placed nipples on the back (191D), at the head; and a
gray IIB bottlelike jar (25T) between the other two vessels. There was no
jewelry.
We now return to the other burials of this period, i.e., those not distin-
guished by stratigraphy as either early or late in deposition. Since this period
is relatively well known from the Hasanlu excavations, only some of the buri-
als need to be presented here. They are representative of the whole corpus:
b9a, 14: Mature adult, flexed on back, N-S, head S; in damaged brick tomb
(Figure 31). Furniture (Figure 32): three plain penannular bronze armlets
on R arm (306T, 365T, 372P); four bronze penannular rings by hands (463T,
594P); a bronze torque with curled ends at neck (1040T); a stone button
(591T); a necklace of plain round paste beads (327T). Beneath the skull, sixty-
six astragals and two bronze buttons (1005T). At the feet, a gray bowl (D)
placed under an orange IIB spouted vessel (336T); a few feet from the face,
an orange IIB vessel with two animal-head lugs (401T). Sheep/goat bones at
the knees.
b9b, 13: A stone tomb (Figures 33, 34, top): only a few bones extant;
apparently N-S. Furniture: outside the entrance slab, an orange IIB jar (89T).
Inside, two plain bronze bracelets with overlapping ends (370T), a red-
slipped spouted vessel (850P), an orange IA jar (I70D), a gray matt carinated
bowl (874D), and a coarse, disintegrated vessel. This was the poorest of the
stone tombs.
b10a, 6: A stone tomb with hard-packed earth floor (Figure 35; Mus-
carella 1968, p. 191, figs. 13, 14). The few bones suggest a N-S orientation of an
adult. Furniture (Figures 36, 37): fifty-three objects; this was one of the rich-
est burials in the cemetery. Objects probably belonged to a male warrior. Fif-
teen vessels (not sixteen as Muscarella 1968, p. 189) were outside the closing
slab: buff: 81D, a fragment of a wide-mouthed pot with an oblique spout and
no handle; also a disintegrated vessel; buff, IIB: 29D, a broken carinated jar;
332D, a jar; orange IIB: 31P, a carinated jar; 32P, a sharply carinated jar with
incisions on the upper body; 33D, a brokenjar; 34T, a spouted vessel; 35T,
a sharply carinated deep bowl with a handle, now missing; 48T, a spouted
vessel; red-orange IIC: 40T, a gadroonedjar; red-orange IC: 423P, a spout-
ed vessel; red-slipped: 30T, an asymmetrical jar; gray IC: 46T, a vertically-
bridged spouted vessel, with broken spout; gray IIB: 28P, a gadrooned jar.
Placed among these vessels was the dismembered, incomplete skeleton of
a horse: skull, mandible, humerus, a pair of radii, two cannon bones, two
femera, one tibia, and a third cannon bone. Just outside the tomb, by the
northwest corner, was an iron socketed spear (118P, visible in Muscarella
1968, p. 121, fig. I4). Just inside the entrance of the tomb were five vessels: an
orange matt hydria (248D), two buff carinated bowls (858P, 910D), a small
buff IA jar (169D), and a gray IIB spouted vessel (239T). Two bronze penan-
nular anklets (761P, 762T) at the southern part of the tomb give the feet
210 chapter six
Figure 32.
Figure 33. B9b, burial 13.

212 chapter six
Figure 34.

214 chapter six
Figure 36.
Figure 37.
216 chapter six
Figure 39.

218 chapter six
Figure 43.
Figure 44.
220 chapter six
Figure 45.
Figure 46. B8e, burial 5.

222 chapter six
position. With the pottery at the entrance was a corroded iron point (D),
a corroded iron object with a rounded head and spike (182D), apparently
a mace head with a solid head; also an iron blade (151P), an iron pin with
traces of eight layers of cloth (706P), a thick-knobbed iron pin, also with
traces of cloth (1031T), a plain round penannular bronze bracelet (129T),
and remains of a bronze and iron chain (1054D, compare Figure 44, B8a, 1).
Along the western wall, N-S, were a bronze spiked or star mace head (119P), a
plain bronze bowl (114P), an elaborately decorated flat-band bronze bracelet
(112T), and a plain concave-sided band bronze bracelet with overlapping
ends (113T), along with two plain round iron bracelets with overlapping ends
(417P), plus a plain broken round iron (134T) and a plain penannular iron
bracelet (120T), for a total of six. There were also clusters of plain iron (426T)
and clusters of penannular bronze rings (593D), a bronze, two-piece, jointed,
horse bit with a solid ring (1026P), and an iron fragment of another horse bit
(69T), an iron shaft-hole ax fragment (1033D), a broken iron archers ring
(195T), a bronze needle (470D), a bronze boss (150P), a bronze stud (1007T),
a limestone disc (64P), and a pin consisting of an iron hooked-top set into
a bone button and attached to a reed, with traces of thread (755T, compare
Figure 47, 756, 757), two bone awls (222T, 223P).
There were also many beads (994T): carneliana, n-r; pastest; chalky
materials; amberb; glasse-g (the latter blue and yellow)u;
cowrie shellh; Egyptian blueil; cast antimonyc; bronzem.
b10a, 12: Male, flexed on R side, N-S, head S; in brick tomb (Figure 38).
Furniture (Figure 39): a plain round bronze and a plain round iron bracelet,
both with tapered overlapping ends (123P, 124T), by wrists; carnelian, frit,
and paste beads (117D); a beaded cast bronze torque with hooked ends, at
neck (115T); a bronze socketed spear resting along L side of head, point up
(125T); shaft would have rested along side of the body. By the knees, two
gray burnished spouted vessels (835T, 846P); by thighs, an orange smoothed
carinated bowl (873D); by feet, an orange burnished carinated jar (965D).
b10a, 13: Female, adult, inhumation. Bones were found disarticulated,
probably representing a secondary burial (Figure 40). Furniture: a bronze
stud (925P) was found inside the skull cavity; bronze hemispherical beads
with a loop, corroded together in sets of three (923P), and carnelian and frit
plain beads, all in the fill (436T); a simple iron ring was under the skull (130T).
Also under the skull was an orange IIA jar (98P); other vessels included a gray
IIC spouted vessel (261T), an orange IIA, and two buff IIA jars (192D, 193D,
94P). Sheep/goat bones.
b10a, 15: Stone tomb; part of the floor covered with stone slabs (Fig-
ures 41, 42). Bones disintegrated; N-S orientation, head N. Furniture: three
corroded knob-headed iron pins at N of chamber (146P, 147P, 186T); two

more of same type at SW corner (196T, 197P). In the fill, two bronze figure-
eight hairrings (earrings?) (208D, 212P), a collection of beads (Figure 51,
815P): acarnelian; b, c, d, i, jpaste; ebronze; fEgyptian blue; g, h
antimony; also five plain iron rings (217T, 218D), two bronze rings (189P,
216P), some bronze studs (899T); a small iron hooked pin (1030D), and three
iron archers rings (207T, 209P, 1028D). On floor, fifteen pottery vessels: two
buff IA hydriai (915D, 916D), a large orange IIB spouted vessel with crows
feet decoration in relief at the rear (268P), two buff IA carinated bowls
(859T, 860P, both containing sheep/goat bones), four jars: two buff IIB (178D,
266D), one buff IA (172D), and one orange IIB (254T), and an orange IIB cup
(232T). In addition, there was a red-slipped, IIB gourd-shaped vessel pierced
by two holes at one side (226M; Muscarella 1968, pp. 189190, fig. 11). There
were also two gray burnished spouted vessels (839T, 906D) and two gray jars,
one burnished (251D), one smoothed (962D).
b10b, 8: Female, old adult, with arthritic lipping of the vertebrae; on
back, N-S, head N; in brick tomb, partly left in the balk (Muscarella 1968,
p. 189, fig. 2). R arm was flexed across the body, L arm was bent back to its
own shoulder, a feature found in Period III. Furniture (Figure 43): a bronze
knobbed pin at L shoulder (127T) and R (149P) (similar to Figure 27, 375);
a flat-band iron ring with tapering ends (133T) on L hand; a bronze needle
below R shoulder (135T); an iron ring with cloth traces (162P) on floor; and
round carnelian, paste, and bronze beads (442T) inside the bowl 227. By the
face were a deep gray IIB bowl with animal-head protome handles (227T,
compare Muscarella 1968, p. 191, fig. 12, left, from B10b, 16), and a gray IIB
carinated jar (250T); by the knees was a buff matt hydria (918T). Sheep/goat
bones on floor.
b8a, 1: A stone tomb; noted sticking out from the eroded north slope
of the mound; excavation was conducted as a salvage operation. 1 had a
neatly laid stone floor and preserved two skulls along with a few other bones;
body positions could not be reconstructed (Figure 44, bottom). Furniture
(Figures 44, 45): two plain round bronze bracelets with overlapping ends
(1012P, 602P), a flattened bronze penannular bracelet (1020T), and an iron
one (D); three iron pins (701D, 709D, 1034actually 125D); three bronze
rings around a finger bone (1013D); bronze figure-eight hairrings (1024D); a
bronze needle (1016D); a bronze tack (1011T); two thin bronze strips attached
to iron loops (1044P, from a chain?); fragments of a chain (1041T, a, iron; b,
bronze) found next to a corroded iron bracelet (not catalogued); three iron
blades with curved tips (623T, 624P, 626T), two of these blades (624, 626) had
cloth remains, one (626) had wood remains on the hilt; an iron dagger with a
224 chapter six
splayed pommel and a straight grip, and with wood fragments of sheathing
and hilt insets evident (1046T); a bronze chain (1034P) next to a large iron
object (found exploded) that might have been a staffor batonit seems
to be too big for a pin, which it resembles (1032D); a corroded iron point,
possibly a large pin, with a bronze chain (1035T); a bone cosmetic container
open at both ends and decorated with incised circles with a dot, empty
(1047P); and under bowl 872 a handful of beads (1049T): a, bamber; c-
fcolored glass (yellow, brown, white, and black); gglazed material; also
beads of Egyptian blue and carnelian. In the fill of the chamber were found
two bronze needles (1017P); four plain bronze penannular rings (1018T), and
a fragment of a bronze coil (1002P). A total often vessels were placed in the
tomb, one of which had disintegrated: a gray burnished tripod bowl (894P),
a gray burnished hydria (917T), a gray matt jar (824D); a burnished orange
carinated bowl (872P), and buff matt vessels: 821P, 822D, 825T, 895P; 841P is
burnished.
b8e, 5: A stone tomb with stone floor, containing the scanty remains of
two individuals, male and female (I). Body positions, not clear, seem to have
been N-S. Outside the tombs western entrance or closing slab was a large
pile of pottery partly covered by a broken pithos (Figure 46). Here were
twenty-three vessels and under them were two skeletons, a female young
adult and a child (II), separated from the stone tomb by a mud brick wall
of one course. Wall ran N-S, paralleling the stone tomb, interrupted where
it touched the western wall; its total excavated length (measuring the area
occupied by the stone tomb section) 2.40 m. Both ends continue into the
unexcavated balks, so we do not know the total length. The female and child
were placed head to feet in a line, separated by about 45 cm.; they were flexed
on their R sides, N-S, heads S, facing the wall, E. Contents of stone tomb (I)
(Figure 47): there were five vessels: two bowls, one gray matt (863P), one
gray burnished (864T); one gray IIB spouted vessel (278M); one large buff IIB
jar with three nipples in triangular form on the sides (1055D); and one red-
slipped broken jar with narrow neck and two handles (806D). Body furniture
(Figure 48): about twenty plain bronze and iron rings (like 489D, 218D); a
spiral bronze ring (490D), and a flat-band iron ring attached to a round one
(188P); four figure-eight hairrings (219P, 488D, 1025D), a bronze coil (769T);
four bronze penannular bracelets (432P, 455P, 456P); two iron penannular
bracelets or anklets (220P, 710T); one iron needle (1021D). Also, three iron
pins with looped heads (486T, 714T, 715P); an iron pin with a knob head
(416P), and two iron pins with ribbed heads (699D, 700T); three plain iron
pins with blunt heads (407T, 409P, 410P); also two iron-reed pin-hooks, with
a bone collar (756T, 757P, of the same type as Figure 36, 755). There were also
Figure 47.
226 chapter six
Figure 48.
two iron archers rings (483T, 485P); two socketed iron spear heads (650T,
702D broken); and an iron knife with a curved tip (704P). Each of the two
outside skeletons had associated grave goods. The northernmost one, the
child, had a pottery jar that disintegrated, an iron bracelet with overlapping
ends (412T), two iron bracelets of the same type found together (625P), and
a plain bronze bracelet (318T). The female had two buff IA bowls (865P,
875D) containing sheep/goat bones; also two bronze penannular bracelets
on one arm (367T), and one on the other arm (368P); also a dark-stone,
pear-shaped mace head (1019P). Vessels found outside the stone tomb over
the skeletons of burial II (Figure 47): gray IIC: spouted vessels (403P, 848P,
845T, 857D, 849P, 333T), bottle (790P), cup (805P); gray weathered: hydria
(919D); gray IE: carinated jar (963D); orange IIC: spouted vessel (840P); buff:
hydria (912D); buff IB: hydria (911D); buff IIA: jar (964D); buff IA: jars (812T,
1058D), cups (800D, 807D), carinated bowl (876P); red-slipped: jar (809D).
A few objects were inadvertently not recorded as specifically coming from
the burials in I or II and are listed here together: an orange jar (262D), a buff
carinated bowl (D), and a disintegrated vessel. Also, a bronze tack or stud
(707T), a stone ax or pestle (1056D), a bronze coil (769T), an obsidian blade
(705T), and many beads: 397Ppaste; 713Pbronze; 708Tglass; 997aP
shell; c, jpaste; g, a spacer beadpaste; hstone; ibone or shell; 998
paste.
The problem of the relationship between the stone tomb (I) and the
burials outside (II) remains to be discussed. What is clear is that the twenty-
three vessels were placed partly over both the west wall of the tomb and the
brick wall, and the skeletons of burial II. Therefore, both burials I and II were
in place and were exposed at the time when the pottery was deposited as a
final act. Yet, what is unclear is whether or not one of the burials was already
in existence before the second was deposited, i.e., whether the diggers of
the second burial pit inadvertently disturbed the earlier burial, or, whether
both burials were deposited simultaneously, with different treatment given
the respective bodies. If we prefer the first suggestion we can assume that
the burials in II existed first, and that it was accidentally encountered by the
stone tomb builders, who, upon completion of their funeral tasks, piled the
many vessels over both burials as a pious gesture. If we prefer the second
suggestion we must assume a unique occurrence at Dinkha: the fact that
at one time four people were buried, two in a closed stone tomb, and two
outside. It should be noted that the brick wall was only one course high and
its length very long, features not encountered with typical brick tombs at
Dinkha. Moreover, the first suggestion implies that when the earlier burial
was encountered, instead of recovering it and going elsewhere, the stone
228 chapter six
tomb builders completely uncovered the bodies. A third possibility presents

itself at this point, namely that the whole unit could represent a family
vault, the bodies placed there at different times, and that the pottery deposit
occurred at the time of the final burial. This suggestion would explain the
uncovering of both burials I and II. I prefer to leave the interpretation open
rather than force a conclusion, but I lean toward the suggestions of a family
vault or simultaneous deposition.
Urn Burials
Nineteen urn burials were excavated at Dinkha Tepe. None of these could be
attributed to the Dinkha III period either by low position or by grave goods,
but it is not impossible that a very few might have belonged to that time.
Most, if not necessarily all, were obviously laid down in the Iron II period
as they were usually found high in the Dinkha II fill; in a few cases they had
characteristic pottery associated with the urn.
The burial urns were either buff matt storage vessels or large cooking pots,
with both wide and narrow mouths (Figure 49: 108D, 111D, 284D). One urn
was blackened on the outside and inside, no doubt from use; often the urns
were broken or incomplete. Two urns were buff hydriai, and in another case
the top of a large pithos was used as an urn; often the mouths of the urns
were covered with large sherds. Urns were usually placed on their sides, but
a few were found upright, or upside down. Those on their side were oriented
N-S or E-W, following the same pattern practiced in the burials. In most
instances few or no bones were recovered from the earth fill inside the urns.
When the bones were recognizable they were usually those of infants, but
in one case (B9b, 2), an adult tooth was found in an urn (compare Stein
1940, p. 374; see also pp. 397, 400). Seven urns had pottery placed outside,
and four, two of them with gifts outside the urn also, had pottery and jewelry
inside.
A sampling of the urn burials:
b9a, 3: The buff urn (D), tilted up, was covered with a broken buff matt
bowl (909D) (Figure 50). Close to the mouth of the urn were a red-orange
IIC spouted vessel decorated with a crescent and two nipples on both sides
(259T), two small gray IIB jars (79T, 80T), and a large buff jar (102D). Inside
the urn were a plain bronze band penannular bracelet (Figure 52, 620P),
two plain bronze (464T, 467P), and one iron ring (425D), thirty-nine bronze
(386T), stone, paste, and shell beads (376P), and a clay button (D). This was
the richest urn burial excavated.
Figure 49. Urn burials.
b10a, 2: (Figure 49): The buff IA urn (108D) was lying on its side with the
mouth to SE; the mouth was broken away. Outside was an orange IIB two-
handled jar or flask that had a short upright spout (10T).
b10a, 3: (Figure 49): The buff IA urn (284D) was placed with the mouth
up and sealed by sherds. An infants tooth was found inside. Near the urn
was a red-slipped carinated jar (905D).
b10a, 5: (Figures 49, 51): The buff urn (D) was lying on its side, roughly
E-W, mouth E; a large sherd sealed the mouth. To one side were an orange
IIB spouted vessel (38T), an orange IIB carinated jar (36T), and an orange-
brown IB miniature asymmetrical jar (15T).
230 chapter six

test trench 111, 1: (Figure 49): the buff-yellow IB (281D) urn was on its
side, facing NW; it contained infants bones and a bronze ring (D); outside
was a gray jar (D) and an orange IIB footed bowl with a hole below the rim
(231P).
Dinkha IIIII
Much has been written about the Iron II period in western Iran, so it would
serve no useful function to repeat that information here except in those
instances where it relates to Dinkha Tepe. That there was cultural continuity
in the Iron Age, that Iron II followed Iron I peacefully and without any
observable interruption from outside forces, is well supported by the Dinkha
excavation. The most obvious evidence is that the Dinkha II burials were
deposited in the same cemetery area as the earlier ones, that both simple
232 chapter six
inhumation and the use of brick tombs continued, that the same body
positions and orientations continued, and that earlier customs, such as the
extended burial and the placement of an arm bent back, touching its own
shoulder, were not forgotten, although rarely practiced. In addition, the
same types of grave goods, pottery, jewelry, weapons, and food continued
to be placed with the dead. Not least in significance is the fact that the
same buff and gray pottery were produced in both periods. But it is of some
interest to note that whereas in Dinkha III gray vessels were more than twice
as common as the buff wares, in Dinkha II the ratio is strikingly reversed and
buff vessels were nearly three times as common as gray ones (see pp. 38, 59).
Obviously, as is to be expected with a dynamic culture that existed for such
a long time, and as is the case at other Iron Age sites, the pottery shapes
changedthe worm bowl and pedestal-base goblet disappearand the
variety of shapes increased. Yet even within the changing pottery repertory
we are able to observe a continuity between Dinkha III and II: the ubiquitous
spouted vessels, the basket-handled teapot, the carinated bowls, and many
of the jar types.
Only one vessel from the corpus of Dinkha Iron II pottery might be singled
out as a possible import: the gourd-shaped red-slipped vessel from B10a,
15. Three other vessels of this type, two red and one gray, were found in
Hasanlu IV (Rad, Hakemi 1950, pp. 5960). Another, exactly the same in all
details, and also red, was seen in the Rezaiyeh market by Kleiss, who related
it, incorrectly, I think, to seventh century Urartian ceramics (Kleiss 1971, p. 71,
fig. 22, pl. 10:3, left). If these vessels were not locally made at Hasanlu or
Dinkha, we do not yet know their source.
Among the similar kinds of jewelry placed in the graves of both periods,
bracelets, anklets, rings, necklaces, pins, we might single out torques for
special mention. Of thirty-four Period III burials, seven contained a torque
(see note 5); of fifty Period II burials (not counting urns), four contained
torques, a drop from about one-fifth to about one-twelfth. Thus, although
still used, fewer people wore themat least to their graves. It seems that
men, women, and children wore torques, although the evidence for this is
clearer in the earlier burials.
There is one example of a disarticulated burial (B10a, 13) in Period II,
none in the earlier period, but whether this is culturally significant or merely
an occurrence reflecting a local situation is not known. Also, as mentioned
above, in Period II four burials contained two individuals each, a feature not
encountered in the earlier period.
And one burial in Period II (B10a, 6) was associated with a fragmen-
tary skeleton of a horse placed outside the tomb. Within the tomb, it will
be recalled, was found a complete bronze horse bit and an iron fragment of
another. No other burial at Dinkha yielded either a horse bit or a horse skele-
ton. It is therefore not rash to conclude that the horse probably belonged
to the occupant of the tomb and was dispatched as part of the burial cere-
mony. But where were the other bones of the skeleton? Does the fragmen-
tary nature of the skeleton suggest that the horse was in fact an ordinary
animal merely meant for the funeral feast, and that the bones by the tomb
represent the dead mans share? It seems to me that the juxtaposition of the
horse bit within the tomb and the horse skeleton outside is not fortuitous,
and that the horse did have some special relationship to the tombs occu-
pant. At the same time it would appear that the horse was eaten and that
some joints were kept for the mourners. I suggest, therefore, that the horse
belonged to the occupant of the tomb and also that the survivors ate it: both
ideas need not be mutually exclusive. Keeping this in mind, I believe that we
may correctly refer to the existence of a horse burial at Dinkha, as opposed
to the idea that the bones represent simply a food deposit. What remains
puzzling is the uniqueness of the occurrence of horse bones at Dinkha, even
if one disagrees with the conclusion presented here and believes the bones
are food.15
Horse burials associated with a human burial occur at Hasanlu in a
unique grave excavated in 1947; to date no other example has been found
even though the cemetery area has been extensively excavated (Ghirshman
1964, pp. 2427, 99, fig. 131; Dyson 1965, pp. 208212). Unfortunately, the
contents of this grave have never been identified and published and the date
is not known, which makes it impossible to bring it into a discussion of Iron
Age horse burials (Muscarella 1968, p. 192).16
15 S. Piggott, Heads and Hoofs, Antiquity 3 (1962) pp. 110118, summarizes information
concerned with the burial of a horses head and feet, presumably along with the hide. The
Dinkha horse burial is of a different type. The description of Scythian horse sacrifices given
by Herodotus, IV: 62, 72, does not reflect light on the customs at Dinkha.
16 When discussing the horse burial at Hasanlu, Dyson (1965, p. 211, and also Young 1967,
p. 33) stated that a Scythian-like cheek piece found at Hasanlu came from Period IV (see
also Dyson 1964c, p. 372, fig. 3). This is an error; the piece came from a Period III context.
The error was repeated by M. van Loon in JNES 29 (1970), p. 69, and by P.R.S. Moorey 1971,
p. 109, and in Iran IX (1971) p. 121. Horse burials occur in the Hittite period, K. Bittel, Die
Hethitischen Grabfunde von Osmankayesi (Berlin, 1958) (only skulls and leg bones pp. 16, 24,
63, 65, 72, 73), and in the Mycenaean period, E. Vermeule, Greece in the Bronze Age (Chicago,
1965) pp. 298299, P. Kabbadias, Proistoriki Archaiologia (Athens, 1909) p. 290; A.J.B. Wace,
Chamber Tombs at Mycenae, Archaeologia 87 (1932) p. 14; AA 1930 p. 170 for a buried
terracotta horse; see also the Iliad XXIII, line 170 et seq. Horse burials also occur in Gaza, in the
second quarter of the second millennium bc, F. Petrie, Ancient Gaza I (London, 1931) pp. 45,
pl. lvii: Hyksos? Horse burials occur in Phrygia at Gordion and Ankara, see R.S. Young The
234 chapter six
At Godin Tepe a complete horse skeleton was found in association with

a Period III, Bronze Age, tomb (Young 1969, pp. 1920, fig. 27, pl. xvi). This
seems to be the earliest example known in Iran of a horse burial. In separate
graves at Marlik, that is, in graves not associated with human burials, horses
teeth with bits in situ were found (Negahban 1964, pp. 15, 16). This type of
individual burial of horses heads with bits seems also to have occurred in
Luristan (Moorey 1971, p. 103). We have no information at present about the
dates of these Marlik burials.
A complete horse with artifacts, including a horse bit, was buried in an
individual grave, not associated with a human, at Baba Jan, sometime in the
eighth or seventh century bc (i.e., post Iron II in northern terminology; Goff
Meade 1969, pp. 123126; Muscarella 1968, p. 192).
Summarizing all this information, one notes that horse burials are doc-
umented in Iran in the Bronze Age (Godin Tepe), in Iron II (Dinkha), and
later (Baba Jan).
The major technological change that occurs in Iron II, recognized not
only at Dinkha and Hasanlu, but also at all the other sites, is the use of iron
alongside bronze. (In fact, one may state in parenthesis, if it would not cause
confusion about continuity of culture and upset established terminology,
one should think of the Iron I period as the Late Bronze Age and the Iron II
period actually as Iron I.)
Hasanlu IV and Dinkha II
Dinkha II and Hasanlu IV continued the close relationship existing from

the Iron I period and there must have been sustained communication and
exchanges. For, aside from the basic architectural features, such as the use
Nomadic Impact: Gordion, Dark Ages and Nomads (Istanbul, 1964) pp. 5556, and T. zg,
Belleten XI (1947) p. 80, all from post-destruction tombs. Of apparently contemporary date
are three horses buried in a stone-lined tomb at Norsuntepe, H. Hauptmann, Norsun-tepe,
1970, Anatolian Studies XXI (1971) p. 20. For horse burials in the Caucasus see F. Hancar, Das
Pferd in prhistorischer und frhen historischen Zeit (Munich, 1955) pp. 180ff. Horse burials
also occur in Cyprus in the eighth and seventh centuries bc, V. Karageorghis, Excavations in
the Necropolis of Salamis (Cyprus, 1967) Tombs 2, 3, 47. A. Hakemi, Kalaruz, Archaeologia
Viva 1 (1968) p. 65, mentions horse burials but gives no details. That such burials continued
in later times in Iran is attested by horse and human bones found together in a first century bc
tomb at Shahr-i Qumis excavated by John Hansman and David Stronach in 1967 (Journal of the
Royal Asiatic Society [1970] I, pp. 4148). An account of horse burials in the Altai region is given
in S.I. Rudenko, Frozen Tombs from Siberia (University of California, 1970). The widespread
occurrence of horse burials over a wide chronological and geographical range indicates that
no one ethnic group had a monopoly on the practice.
of posts in the Dinkha II building, and the juxtaposition of jube and pave-
ment in Structure A, features shared by both cities, practically every pot-
tery shape used by the people at Dinkha II was used by the people of
Hasanlu IV: spouted vessels, plain, or decorated with crescents, nipples, or
animals in relief, or decorated with flutings, ridges, crows feet, or with an
animal head at the handle, and sometimes with a vertically bridged spout;
hydriae, which seem to occur only at Hasanlu and Dinkha; basket-handled
teapots; carinated bowls; deep bowls with animal-head protome handles;
the (imported?) pear-shaped gourd; the many jar types; knobbed vertical
loop handles; asymmetrical handmade bowls; gadrooned jars (Dyson 1964a,
figs. 118121; 1965, fig. 13; Young 1965, figs. 6, 7; vanden Berghe 1959, figs. 144
146; Stein 1940, fig. 109, pls. xxiv, xxx, xxxi; Rad, Hakemi 1950, pp. 5960,
Burney, Lang 1972, p. 125).
Equally shared are the many metal and miscellaneous objects, such as
jewelry and weapons: pear-shaped stone mace heads and metal-spiked or
star maces are very common at Hasanlu (Rad, Hakemi 1950, fig. 78b; Dyson
1960, fig. on p. 128); so were iron knives with curved tips, and iron socketed
spears, found there by the hundreds (compare Moorey 1971, pp. 8890;
compare no. 87, to my Figure 26, B9a, 9, 221). Bronze and iron archers
rings were excavated in many Hasanlu burials (Stein 1940, pl. xxv, 2; seven
iron examples were found at Dinkha) and two bone-antler axes (exactly
the same type as in Figure 26, B9a, 9, 1042) were found at Hasanlu, one
in a burial. To my knowledge only one daggeractually the hilt alone is
preservedexactly paralleling the sole example from Dinkha (Figure 45,
B8a, 1, 1046) comes from Hasanlu; but another similar example was also
found there (Dyson 1964a, p. 41, fig. 2:2, pl. ix, 2; see also Moorey 1971, pp. 70
71).
Plain, jointed horse bits of the same type as Figure 36, B10a, 6, 1026, as
well as twisted and elaborate examples occur at Hasanlu in bronze and iron
(Ghirshman 1939, pl. cc: 17; 1964, fig. 338, left; compare also fig. 338, right).17
The iron and bronze chains from B10a, 6 and B8a, 1 (Figure 45, 1034,
1035, 1041) may have been originally attached to pins as was the case with the
many lion pins from Hasanlu (Dyson 1964c, p. 374, figs. 9, 12) and at Haftavan
(Burney 1970, fig. 7, middle). Chains occur at Hasanlu not only in connection
17 The number of horse bits found at Hasanlu gives evidence for the use of cavalry and
perhaps chariotry there, a fact corroborated by the scenes represented on ivories found at
the site, Muscarella 1966, figs. 11, 12 (but no bits are depicted on these horses); see also fig. 10.
It is not possible to make any comments about the extent of cavalry and chariotry at Dinkha
from the two examplesone a fragmentfound in one tomb.
236 chapter six
with lion pins, but individually (although they too may have been connected
to other pins), and in a Period III context, attached to a fibula.18
It was mentioned before that torques were known at Dinkha III and II,
and that they occur in small quantities at Hasanlu IV. Bronze and iron plain
round rings, single and doubled, and flat-band rings are common at Hasanlu.
Several plain band bracelets, some with concave sides (Figure 36, B10a, 6,
113), and several examples of the elaborately incised band types (Figure 36,
B10a, 6, 112), were at home at Hasanlu. The dead at Hasanlu were also
dressed, and wore anklets, plain loop and figure-eight hairrings, and pins,
of exactly the same types as those from Dinkha; they were also furnished
with needles (Stein 1940, p. 401).19
Literally scores of thousands of beads of all typical materials, including
antimony, amber,20 and Egyptian blue, were found in the graves and on
the citadel at Hasanlu. Astragals, polished from use, were found at both
sites; Hasanlu produced some that were pierced. We may assume that the
same games were played at both cities, which is not surprising inasmuch
as knucklebone games have a long history in the ancient Near East, and
in modern history as well.21 Another type of bone object from Dinkha, the
18 Compare also M. Mallowan, Nimrud and Its Remains II (New York, 1966) p. 114, fig. 58.
19 Many of the figures on the Hasanlu ivories wear bracelets, and so do those represented
on the gold bowl; the nude female figure there also wears anklets.
20 It does not yet seem possible to be certain about specific proveniences of ancient
amber: Curt W. Beck, Analysis and Proveniences of Minoan and Mycenaean Amber, Greek
Roman and Byzantine Studies 7, 3 (1966) pp. 191211.
21 Our workers always asked for discarded astragals from the ancient burials to give to
their children; see also C.L. Woolley in LAAA 26 (1930) p. 20, note 1, where it is reported
that astragals were placed in modern childrens graves. Their occurrence in ancient times
is widespread, as the following incomplete listing makes clear: in Iran they are reported,
besides those from Hasanlu and Dinkha, from Geoy Tepe, Burton-Brown 1951, p. 175, note 15,
pl. xxii, A Period; from Sialk B, Ghirshman 1939, p. 245, pl. lxxviii; from Ghalekuti, N. Egami,
et al., Dailaman I (Tokyo, 1965) pl. xlviii, no. 28; from early Susa, J. de Morgan, MMA en Iran
XXIX (Paris, 1943) pp. 46 ff. In Anatolia they occur early at both Hacilar and atal Hyk,
J. Mellaart, Anatolia Before 4000 BC, CAH Fascicle 20 (1964) pp. 10, 14; at Troy, H. Schliemann
Ilios (New York, 1881) pp. 263, 426; at Alishar, E.F. Schmidt and H.H. von der Osten, in OIC XIX
(Chicago, 1932), p. 274, fig. 374; OIP XX (Chicago, 1933) pp. 8283, fig. 129; OIP XXIX (Chicago,
1937) p. 433, fig. 488; OIP XXX (Chicago, 1937) pp. 105, 174175, figs. 101, 196 (late); at Bogazky,
MDOG 72 (1933) p. 77, fig. 12; R.M. Boehmer, Die Kleinfunde von Bogazky (Berlin, 1972) pp. 35,
181, 203. They were very common in Phrygian Gordion: a large vase filled with astragals was
found in one of the burnt buildings, and many were found on the floor of another, R.S. Young
in AJA 61 (1957) pp. 321, 327; 446 astragals were found in Tumulus P, a childs tomb, ibid.,
p. 327. In North Syria they occur at Zincirli, F. von Luschan and W. Andrae, Die Kleinfunde von
Sendschirli V (Berlin, 1943) pp. 122124, fig. 173, pl. 59: p, q; at Hama, P.J. Riis, Les Cimetires
Crmation (Copenhagen, 1948) pp. 30, 35, fig. 22, p. 176; at Carchemish, C.L. Woolley, LAAA
26 (1939) pp. 2021, 23 f.; note also the relief there with children playing an astragal game,
incised cosmetic container from B8a, 1 (Figure 45, 1047), has many relations
at Hasanlu (Stein 1940, pl. xxv, 6; Ghirshman 1939, pl. c, 24; see also Dyson
1964c, figs. 1417).
In short, the two sites shared a common culture. This conclusion is not
contradicted by the fact that there were some differences between the
sites, some traits that were not shared in common. For example, burials at
Hasanlu continued to be simple inhumations, while at Dinkha, alongside
inhumation, the earlier use of brick tombs continued, augmented by the
innovation of stone chamber tombs; and at Hasanlu only a few urn burials of
undetermined date have been found (compare Stein 1940, pp. 397, 400, date
not clear). Certain pottery types, very much in evidence at Hasanlu, do not
occur at Dinkha: tripod stands for supporting spouted vessels, which were
found in many Hasanlu burials, and spouted vessels with animals sculpted
on the spout, or vessels with an animal at the handle (Dyson 1968b, figs. 118,
121; vanden Berghe 1959, pl. 145, c-e); bowls with tab handles, solid and
looped, sometimes with animal-head protomes on the body (Dyson 1964c,
fig. 13; Young 1965, figs. 6:3; 7:3; Boehmer 1967, p. 580, fig. 7); and vessels on tall
hollow stands (Dyson 1964a, figs. 4:7, 9, 10, 11). Nor do we have any evidence
at Dinkha for the fine wares with polished gray surfaces, and for glazed wares
(Young 1965, p. 55).
Lion pins, metal bells, belts, animal figurines, armor, metal and pottery
rhyta, not to mention ivories and vessels made of precious metals, were
not found at Dinkha. But it must be stressed that many of the Hasanlu
objects mentioned come from the destroyed citadel, whereas at Dinkha we
are dealing with a cemetery alone and have only the evidence from material
placed in burials. However, from the sophisticated and massive architecture
preserved at Hasanlu, and from the vast quantity of material remains, both
of local and of foreign manufacture (Muscarella 1971a, pp. 263265), there
can be no doubt that Hasanlu was culturally and economically the richer
E. Akurgal, The Art of the Hittites (New York, 1962) fig. 122. For Assyria see A. Haller, Die Grber
und Grfte von Assur (Berlin, 1954) pp. 18, 2122, 103. For Palestine, see for Lachish, O. Tufnell,
Lachisch II (London, 1940) p. 194; for Ugarit, C.F.A. Schaeffer, Ugaritica IV (Paris, 1962) pp. 80
82, 103105, figs. 64, 65. In Mesopotamia we find them at Tepe Gawra, E. Speiser, Excavations
at Tepe Gawra (Philadelphia, 1935) p. 33; Nuzi, Starr 1939, pp. 378379, 414, 450, and Vol. II,
pl. 117, n; at Kish, in the Field Museum of Natural History in Chicago. They also were used
in Egypt, H. Schfer, Aegyptische Kunst (Berlin, 1913) fig. 122, and Nora Scott, BMMA, Spring
1973, fig. 39. In the West they were common from Bronze Age to Roman times at too many
sites to mention here. See, for example, L. Deubner, Zum Astragalspiel, AA 1929, pp. 272282;
Pauly-Wisowa Supplement, IV (1924) Astragalomanteia, pp. 5155; R. Hampe, Die Stele aus
Pharsalos in Louvre, Winckelmannsprogram der Arch. Gesellschaft zu Berlin (1951); G. Bass,
Cape Gelidonya: A Bronze Age Shipwreck (Philadelphia, 1967) p. 133.
238 chapter six
site, perhaps even the main seat of government and trade in the area.
Dinkha, on the other hand, while obviously not poor, was quite clearly a less
important site, perhaps because it was closer to the western border.
The strong cultural connections demonstrated to exist between Hasanlu
and Dinkha over such a long period of time suggest an hypothesis: that both
the Solduz and Ushnu valleys were part of the same ancient state, of which
Hasanlu may have been the major city, with Dinkha one of several provincial
towns (there are still several unexcavated large mounds in Solduz) governed
by a prince or governor. It is also possible that the same language was spoken
at both sites. To be sure, we know nothing about ancient place names or
languages in the area and therefore can go no further than hypothesizing.22
But with respect to the material evidence of the two valleys in the Iron Age,
they must be treated as one cultural region.
Visible from Dinkha Tepe to the east is the still unexcavated Urartian
site of Qalatgah (Figure 19; Muscarella 1971b, pp. 4449). During survey
work conducted by the Hasanlu Project an Urartian inscription, written for
Ishpuini and his son Menua sometime about 810805 bc, was found. This
important inscription dates the entry of the Urartians from the north into
the southern Urmia basin, specifically, into the Gadar and Ushnu valleys.
And it is at this very time,23 as established by independent archaeological
research, that Hasanlu IV was violently destroyed and Dinkha II was ter-
minated, probably by abandonmentfor it is quite clear that the chronol-
ogy of Dinkha II depends completely on that of Hasanlu. Surely, these two
events, the end of the Iron II culture and the entry of the Urartians into
the area are related: the Urartian invasion of the west and south of the
lake is the historical event that aborted the flourishing Iron II culture. A
few years later King Menua alone set up a stele at Tashtepe, about fifty
miles to the east of Dinkha, demonstrating the southeastern limits of the
invasion.
22 As stated in the text, there are unexcavated mounds in the area of Solduz, and future
work might alter the suggestions made here. At present the Solduz valley is inhabited largely
by Turkish-speaking Shia Moslems, the Ushnu valley by Kurdish-speaking Sunni Moslems.
Future archaeologists might not be able to surmise from the remains of their material
culture as represented by house plans, burial customs, and household goods, that they were
two different cultural groups with different languages and histories, and sometimes mutual
hostility.
23 If the Iron II period ended sometime before 800bc, according to possible interpreta-
tions of the MASCA correction dates, then the building of Qalatgah had no direct connection
with the end of Hasanlu and Dinkha, which would presumably have been in ruins.
Dinkha II and other Iron II Sites
We need only present here a few brief comments about Dinkhas material
relationship to contemporary sites, since much has already been written
about this period. The ties between Hasanlu IV and Sialk B, Geoy Tepe
A (in part, for Iron III remains exist there also: Muscarella 1973, p. 72),
Khurvin, Giyan I1 (part), and the Zendan I (part) are well known and have
been discussed often (Young 1965, pp. 6168, 7072; 1967, pp. 2427; Dyson
1965, pp. 197203; Boehmer 1967, pp. 576585; Burney, Lang 1972, pp. 122
126). And because of Dinkhas close relationship to Hasanlu, the same ele-
ments in the discussion obtain for Dinkha. Although pottery has been the
main element referred to in discussing relationships, we might expand
this by including other objects. Thus, at Sialk B several multiple burials
existed, and chains, plain jointed horse bits, flat-band rings, decorated band
bracelets, and torques were placed in burials there (Ghirshman 1939, pls. l,
lvi, lix, lxviii, lxxxv, lxxvii, lxxviii, lv, etc.; see also Young 1967, pp. 7677,
note 28).
At Khurvin, in addition to the typical Iron II vessels, several metal objects
are of interest to us: torques (see above), tweezers, decorated band bracelets,
and plain bracelets with tapered ends (vanden Berghe 1964, pls. iv, v, xi, xii,
xvii, xxii, xxxix, xli, pp. 2930, pl. xlii).
Grave 4 from Tepe Guran should be mentioned again in this context for
it contained bronze vessels of a type found at Sialk B and similar to some at
Hasanlu IV. The sword also found in the tomb indicates, perhaps, a tenth
ninth century dating for the grave rather than nintheighth (Thrane 1964,
pp. 158160, note 6; compare Moorey 1971, p. 21).
A few more Iron II sites may be added to the growing list of Iron II sites
in western Iran. Yanik Tepe is said to have yielded gray wares of Hasanlu IV
type, but no details are yet available (Burney 1964, p. 60). On the western side
of the lake at Haftavan Tepe, we are informed that an Iron II settlement was
partly uncovered. Moreover, part of an extramural cemetery was excavated
and Iron II burials were uncovered. In one was found a red bridged spouted
vessel, but in other burials dating is not so clear-cut (Burney 1970, pp. 165
168, figs. 7, 8:2). Some of the burials had chains attached to pins, and figure-
eight hairrings (earrings?), anklets, bracelets, rings, and beads; there was
also one Mitannian-type seal (Burney 1970, pp. 165168, fig. 7; 1972, pp. 134 ff.,
figs. 8, 9, pl. ivb). These finds collectively could indicate a date close to 800 bc
Until the complete publication of the Marlik material it is not easy to
argue strongly for an Iron II occupation here. Nevertheless, the excavator
(Negahban 1964, p. 38) and others who have seen the material (Dyson 1965,
240 chapter six
chart on p. 11; Young 1967, p. 22, note 69; Burney, Lang 1972, p. 118) agree that
some of the material from the tombs belongs in the early first millennium
bc (Compare Moorey 1971, pp. 2324, who prefers a late second millennium
bc date.) I, too, think there is evidence for an Iron II occupation there on
archaeological and art-historical grounds (Muscarella 1972, pp. 4243).24
Far away to the southeast at Tepe Yahya (III) a fragment of a gray bridged
vessel was found, and we are told that both gray and red wares occur in this
level (Lamberg-Karlovsky 1970, p. 27, pl. xiii). This information could indi-
cate that there was an Iron Age level at Yahya, but based on the published
material perhaps Iron II/III rather than Iron I/II, as suggested by the excava-
tor.
One final point will be presented here, a point already made by Young
(1967, p. 25), that practically all the sites that had Iron I material also had
Iron II material. Which is to say that from an archaeological view the Iron
Age I and II cultures lasted over a large area for a long time, and may
reflect the historical fact that there was a population continuum in much
of western Iran until the early eighth century bc. Of course, Sialk B is the
anomaly here because of its extraordinary painted-ware tradition, and here
alone one might be able to argue against stability (Dyson 1965, pp. 200
201). The isolated Iron I burials at Dalma and Hajji Firuz, and at Godin to
the south, should be kept in mind, but they do not contradict a continued
distribution of the Iron I and II cultures.
Code for Tables
Burial: Sex/Age: Body Positions: Head Faces:

I: inhumation F female B on back F to feet
B: brick tomb M male R on right side
S: stone tomb I infant L on left side; ext. extended
C child F flexed
YA young adult S arm touches own shoulder
MA mature adult
OA old adult
A adult
24 At Klar Dasht a bridged spouted vessel with three small feet was found: H. Samadi, Les
dcouvertes fortuites Klardasht, Garmabak, Emam et Tomadjan (Teheran, 1950) pp. 8, 12, fig. 9.
Table I: Dinkha III Burial Data
Body Head Leg Arm No.
Orien- Posi- Posi- Head Posi- Posi- Ves- Weapons and Animal
Burial Type Sex/Age tation tion tion Faces tion tion sels Jewelry Miscellaneous Bones Comments Text Reference
B9a11 I C N-S B S F L:F 1 Bracelets, beads Poorly preserved
R:S
15 B F/MA N-S B N F F F 3 Bracelets, pins, X? P. 44, Figs. 9, 52
rings, needle, (385, 400)
headband, beads
16 I ?/MA N-S L S W F F 2 Pin, needle
17 I ?/MA N-S L N SE F R:? 3 Torque, bracelets, P. 44, Figs. 10, 11
L:S anklets, ring, pin,
beads, needle
18 I ?/MA N-S B S E F 1 Bracelet, pins, Poorly preserved
ring, beads
19 I C N-S R S E F L:F 4 Bracelet, ring Pp. 4445, Fig. 12
R:S
the iron age at dinkha tepe, iran
22 B C N-S S E L:F 2 Bronze necklace, Poorly preserved;

R:? ring, bracelet tomb has brick
floor
23 I M/MA N-S S 3 Bracelets, ring, Daggers; seal Poorly preserved Pp. 4043, Fig. 6
bone pendant,
beads
241
242
Table I (cont.)
24 I ?/MA N-S L S F F 2 Bracelet, pins, P. 43, Fig. 6;
rings, needle, Muscarella, 1968,
beads fig. 19
25 I M/MA E-W B E NW F 2 Pin, beads Poorly preserved P. 40, Figs. 3, 5
26 I C N-S L N NE F F 3 Torque, bracelets, Bronze bowl P. 43, Fig. 7;
anklets, pins, Muscarella, 1968,
rings, earrings, fig. 16
needles, beads
27 I M/MA N-S B S NE F L:F 2 Bracelet, beads Spear X P. 43, Figs. 7, 8
R:S
chapter six
B9b, 11 B C N-S L N E F F 3 Torque, bracelets, P. 45, Figs. 2, 13;

anklets, earrings, Levine, 1971, p. 40
beads
12 I C N-S B S F F 3 Ring X Poorly preserved Pp. 4546, Fig. 13;
Muscarella, 1968,
figs. 15, 17 right
16 I ?/YA N-S R S E F F 3 Bronze necklace, Dagger P. 46, Figs. 2, 14
bone awl,
bracelet, beads
18 B C N-S B N E F L:ext. 3 Torque, bracelets, Figs. 2, 52 (269T)
R:ext. anklets, pin,
beads
B9b, 20 I 3 Skeleton in balk
21 B F/MA E-W B E Sky F L:F 1 Bracelets, pins, Partly in balk Fig. 2
R:F ring, beads
B10a, 19 I F/YA N-S R N E F R:F 3 Pins, ring, beads
L:S
20 I M/MA N-S B S Sky F L:F 2 Bracelet, ring,
R:S beads
22 I M/MA E-W B E Sky F L:F 2 Bracelet
R:S
23 I E-W B E Sky F F 1 Bracelet, pin
25 I F/YA N-S R S F L:F 1 Bracelet
R:S
B10b, 10 B F/MA N-S L N E F F 3 Pins, earrings, X P. 46, Figs. 15, 16;
ring, beads Muscarella, 1968,
fig. 2
13 B F/MA N-S L N E F F 3 Pins, earrings,

ring, needle
B8e, 7 B 2 Disturbed P. 47, Fig. 17
8 B ?/MA N-S L N 3 Torque, bracelets, Disturbed Fig.52 (433T, 638P)
pins, button
headband, beads
243
244
Table I (cont.)
TT IV 1 I F/YA N-S L N E F F 3 Torque, pins, Fig. 52 (509T)
earrings, ring,
beads
TT VII, 1 I F/YA N-S R N Ext. at 3 Torque, bracelets, Pp. 3940, Fig. 3
sides pin, earrings
2 B M/YA E-W B E Sky Ext. F 2 Bracelet Dagger P. 40, Figs. 3, 4
TT IX 1 I F/A E-W B W Sky Ext. at 2
sides
5 I I N-S R S 2 Bracelets, anklets, Poorly preserved
pins, beads
chapter six
TT XI, 1 I M/A N-S S F F 4 Bracelets, anklets,

pins, ring, beads
Table II: Dinkha II Burial Data

B9a, 4 B F/A N-S B N S F L:ext. 3 Pins, rings,
R:F hairring beads
7 B C N-S R S E F R:ext. 3 Bracelet, rings, Fig. 52 (202T, iron)
L:F beads
8 B 2 Ring Partly in balk
9 B M/MA E-W B W F F at 5 Bracelets, rings, Spear, antler X Pp. 6061,
sides beads ax Figs. 25, 26
10 B N-S B S 1 Bracelet Partly in balk
14 B ?/MA N-S B S E F F 3 Torque, rings, X P. 63, Figs. 31, 32
bracelets, beads
B9b, 1 I I E-W L E S F F 0 Beads Fig. 2
3 I ?/MA N-S N 2 Ring X Disturbed Fig. 2
6 I C N-S L N E F F 0 Pins, bracelet Fig. 2
7 I C N-S L N E F F 2 Stone under chin Fig. 2
9 B C N-S L N E F F 5 Pins, beads, X Disturbed Figs. 2, 33; Levine,
button 1971, p. 40
10 B ?/MA N-S R S N F F 6 Bracelets, beads Spear X Fig. 2
13 S N-S 5 Bracelets No skeleton P. 64, Figs. 2, 33, 34
14 B F/MA N-S R S E F at 4 Beads Fig. 2

sides
15 B F/MA E-W L E S F 1 Bracelets, anklets, Partly in balk Fig. 2
pins, rings
17 B 0 Partly in balk Fig. 2
245
Table II (cont.)
246

19 B ?/A N-S L N E F F 3 Torque, pins, P. 61, Figs. 2, 27
rings, earrings,
beads
B10a, 6 S ?/MA N-S N F 20 Bracelets, anklets, Spear, maces, No Pp. 6467,
pins, rings, beads, ax, blade, skeleton; Figs. 3537;
needle, chains, archers rings, dismem- Muscarella, 1968,
studs horse bits, bered 189ff., figs. 13, 14
bronze bowl horse
skeleton
7 I M/OA N-S R S NE F R:ext. 4 X
L:F
chapter six
11 I F/MA N-S R S E F 1 Bracelet Partly in balk

12 B M N-S R S E F F 4 Torque, bracelets, Spear P. 67, Figs. 38, 39
beads
13 I F/A 5 Ring, stud, beads X Disarticulated P. 67, Fig. 40
14 S N-S N 4 Bracelets, anklets, No skeleton
pins, hairrings,
earrings
15 S N-S N 15 Pins, hairrings, Archers rings X No skeleton P. 67, Figs. 41, 42,
rings, studs 52 (815); Muscar-
ella, 1968, fig. 11
B I N-S L N 3 Torque, bracelet, Few bones extant P. 61, Fig. 28
beads
17 I M/MA N-S B S N L:F 5 Bracelet, beads Partly in balk
R:S
21 B ?/MA N-S N F 2 Partly in balk
24 I ?/MA N-S B N F 8 Disturbed
B10b, 1 I ?/YA E-W R E N F R:ext. 3 P. 63, Fig. 30
L: at
side
2 I C N-S R S NE F F 2 Ring Partly in balk
3 B C N-S R S E F F 3 Skull had a partly P. 59
healed hole
4 I F/MA E-W B E N F F 0 Pin
5 B ?/YA N-S R N NW F F 5 Pins, hairrings, X Fig. 52 (54T, iron;
rings, earrings, 159T, bronze);
studs, beads, Muscarella, 1968,

tweezer figs. 2, 6
6 B M/MA N-S B S NE F R:ext. 3 Bracelet, ring Blade, point Muscarella, 1968,
L:F figs. 25
7 B M/OA N-S R S E F F 3 Bracelet Arthritic lipping P. 59, Muscarella,
of vertebrae 1968, fig. 2
247
248
Table II (cont.)
8 B F/OA N-S B N E R:F 3 Pins, ring, needle, X Arthritic lipping Pp. 59, 6768,
L:S beads of vertebrae; Fig. 43; Muscar-
partly in balk ella, 1968, fig. 2
9 B C E-W L W N F F 3 Pins, hairrings, Muscarella, 1968,
ring, earrings, figs. 2, 7
beads
11 B F/MA N-S B N E Ext. F 2 Pins, rings, beads Archers ring Pp. 6162, Fig. 29
14 B ?/MA N-S L S W F 4 Partly destroyed Muscarella, 1968,
by 9 fig. 2
B10b, 16 B F/MA N-S L N E F F 7 Pins, rings, Fig. 52 (342T, 343P,
chapter six
hairring, needle iron); Muscarella,

1968, figs. 2, 12
B8b, 1 S 10 Bracelets, anklets, Dagger, knives, Two skulls, no Pp. 6871, Figs. 44,
pins, rings, point skeletons 45
cosmetic box
B8d, 1 B F/MA E-W R E N F R:at 6 Mother and child P. 59
I side
L:F
2 B N-S L B E 2 In balk
B8e, 3 B N-S S E F 1 Bone needle Disturbed; in balk
4 B N-S B N F 2 Disturbed
251 S M/A F 5 Bracelets, pins, Spears, knife, Few bones extant; Pp. 7274,
F/A 23 hairrings, rings archers rings two skulls Figs. 4648
511 B C N-S R S F 1 X Two skeletons Pp. 7274,
Figs. 4648
F/YA N-S R S F 2 Bracelets Mace
6 B N-S S E F 3 Beads
G10c, 8 B ?/A N-S S F F 4 Disturbed
Table III: Burials of Undetermined Period

Orien- Body Head Head Arm No. Weapons and
Burial Type Sex/Age tation Position Position Faces Position Vessels Jewelry Miscellaneous Comments Text Reference
B9a, 21 I 0 Bone pin Low in fill; disturbed
B10a, 18 I M/MA E-W B E Sky L:F 1 Partly in balk
R:S
26 I B Bronze dagger Disturbed Fig. 52 (648P)
B10b, 12 I F/MA N-S N W 0 Disturbed
249
250 chapter six
Bibliography
Barnett 1967
R.D. Barnett, The Urartian Cemetery at Igdyr, Anatolian Studies 13 (1967) pp. 153
198.
Boehmer 1965
R.M. Boehmer, Die Grabungen auf den Zendan-i-Suleiman, AA (1965) pp. 746787.
Boehmer 1967
Forschungen am Zendan-i-Suleiman, AA (1967) pp. 573585.
Burney 1962
Charles Burney, Excavations at Yanik Tepe, Azerbaijan, 1961, Second Preliminary
Report, Iraq 24, 2 (1962) pp. 134149.
Burney 1964
Excavations at Yanik Tepe, Azerbaijan, 1962, Third Preliminary Report, Iraq 26, 1
(1964).
Burney 1970
Excavations at Haftavan Tepe 1968: First Preliminary Report, Iran VII (1970) pp.
157171.
Burney 1972
Excavations at Haftavan Tepe 1969: Second Preliminary Report, Iran X (1972)
pp. 127142.
Burney 1973
Excavations at Haftavan Tepe 1971: Third Preliminary Report, Iran XI (1973) pp. 153
172.
Burney, Lang 1972
Charles Burney, Andrew Lang, Peoples of the Hills, (New York, 1972).
Burton-Brown 1951
T. Burton-Brown, Excavations in Azerbaijan, 1948 (London, 1951).
Carter 1965
T.H. Carter, Excavations at Tell al-Rimah, 1964, BASOR 178 (1965) pp. 4069.
Contenau, Ghirshman 1935
G. Contenau, R. Ghirshman, Fouilles de Tep Giyan, (Paris, 1935).
Dyson 1958
Robert H. Dyson, Jr., Iran 1957: Iron Age Hasanlu, Univ. Museum Bulletin 22, 2 (1958)
pp. 2532.
Dyson 1960
Hasanlu and Early Iran, Archaeology 13, 2 (1960) pp. 118129.
Dyson 1962
The Hasanlu Project, Science 135, no. 3504 (1962) pp. 637647.
Dyson 1964a
Notes on Weapons and Chronology in Northern Iran around 1000bc, in Dark Ages
and Nomads, ed. M. Mellink (Istanbul, 1964) pp. 3245.
Dyson 1964b
Sciences Meet in Ancient Hasanlu, Natural History Oct. 1964, pp. 1625.
Dyson 1964c
In the City of the Gold Bowl , ILN, Sept. 12, 1964, pp. 372374.
Dyson 1965
Problems of Protohistoric Iran as seen from Hasanlu, JNES 24, 3 (1965) pp. 193217.
Dyson 1967
Early Cultures of Solduz, Azerbaijan, Survey of Persian Art XIV (Tokyo, 1967) pp.
29512974.
Dyson 1968a
The Archaeological Evidence of the Second Millennium BC on the Persian Plateau,
CAH, fascicle 68 (Cambridge, 1968) pp. 336.
Dyson 1968b
Hasanlu and the Solduz and Ushnu Valleys, Archaeologia Viva 1, 1 (1968) pp. 82101.
Ghirshman 1939
R. Ghirshman, Fouilles de Sialk (Paris, 1939).
Ghirshman 1954
Village perse-achmnide, MMA en Iran XXVI (Paris, 1954).
Ghirshman 1964
The Arts of Ancient Iran (New York, 1964).
Godard 1931
A. Godard, Les Bronzes du Luristan (Paris, 1931).
Goff Meade 1968
C. Goff Meade, Luristan in the First Half of the First Millennium B.C., Iran VI (1968)
pp. 105134.
Goff Meade 1969
Excavations at Baba Jan 1967 , Iran VII (1969) pp. 115130.
Hancar 1934
F. Hancar, Kaukasus-Luristan, ESA 9 (1934) pp. 47112.
Herzfeld 1941
E. Herzfeld, Iran in the Ancient East (London, 1941).
Kleiss 1971
W. Kleiss, Erkundungsfahrten in Iran 1970, AMI 4 (1971) pp. 51111.
Lamberg-Karlovsky 1970
C.C. Lamberg-Karlovsky, Excavations at Tepe Tahya, Iran, 19671969: Progress Report I
(Harvard University, 1970).
252 chapter six
Levine 1971
Louis D. Levine, The Iron Age Revealed, Expedition 13, 34 (1971) pp. 3943.
Moorey 1971
P.R.S. Moorey, A Catalogue of the Persian Bronzes in the Ashmolean Museum (Oxford,
1971).
Muscarella 1966
Oscar White Muscarella, Hasanlu 1964, BMMA Nov. 1966, pp. 121135.
Muscarella 1968
Excavations at Dinkha Tepe, 1966, BMMA Nov. 1968, pp. 187196.
Muscarella 1971a
Hasanlu in the Ninth Century BC and its Relations with other Cultural Centers in
the Near East, AJA 75 (1971) pp. 263266.
Muscarella 1971b
Qalatgah: an Urartian Site in Northwestern Iran, Expedition 13, 34 (1971) pp. 44
49.
Muscarella 1972
A Bronze Vase from Iran and Its Greek Connections, MMA Journal 5 (1972) pp. 25
50.
Muscarella 1973
Excavations at Agrab Tepe, Iran, MMA Journal 8 (1973) pp. 4776.
Negahban 1964
E. Negahban, A Preliminary Report on Marlik Excavation (Teheran, 1964).
Rad, Hakemi 1950
M. Rad, A. Hakemi, The Description and Results of the Scientific Excavations at
Hasanlu, (Teheran, 1950) (in Persian).
Ralph et al. 1973
E.K. Ralph, H.N. Michael, M.C. Han, Radiocarbon Dates and Reality, MASCA News-
letter 9, 1 (Aug., 1973) pp. 118.
Schaeffer 1948
C.F.A. Schaeffer, Stratigraphie Compare et Chronolgie de lAsie Occidentale, (Oxford,
1948).
Schaeffer 1949
Ugaritica II (Paris, 1949).
Schmidt 1970
E.F. Schmidt, Persepolis III, The Royal Tombs and Other Monuments (Chicago, 1970).
Starr 1939
R.F.S. Starr, Nuzi (Cambridge, 1939).
Stein 1940
A. Stein, Old Routes of Western Iran (London, 1940).
Thrane 1964
H. Thrane, Excavations at Tepe Guran, Luristan , Acta Archaeologia XXXIV (1964)
pp. 121133.
Thrane 1965
Archaeological Investigations in Western Iran, Acta Archaeologia XXXV (1965)
pp. 153169.
Thrane 1970
Tepe Guran and the Luristan Bronzes, Archaeology 23 (1970) pp. 2736.
vanden Berghe 1959
L. vanden Berghe, Archologie de lIran Ancien (Leiden, 1959).
vanden Berghe 1964
Le Ncropole de Khurvin (Istanbul, 1964).
Woolley 1955
C.L. Woolley, Alalakh (London, 1955).
Young 1962
T. Cuyler Young, Jr., Taking the History of the Hasanlu Area Back Another Five
Thousand Years, ILN, Nov. 3, 1965, pp. 707709.
Young 1965
A Comparative Ceramic Chronology for Western Iran, 1500500 bc, Iran III (1965)
pp. 5385.
Young 1966
Survey in Western Iran, 1961, JNES XXV (1965) pp. 228239.
Young 1967
The Iranian Migration into the Zagros, Iran V (1967) pp. 1134.
Young 1969
Excavation at Godin Tepe: First Progress Report, Royal Ontario Museum, Art and
Archaeology, Occasional Paper 17 (Toronto, 1969).
chapter seven
WARFARE AT HASANLU IN THE LATE 9TH CENTURY BC*
Warfare in the ancient Near East is abundantly documented by written

and archaeological evidence. The use of force to settle political disputes,
and to validate the role of kings or leaders is not only common but is
glorified in both historical texts and representational art. Excavations at the
site of Hasanlu have produced information about the culture, peaceful or
otherwise, of the people who inhabited the site at the beginning of the first
millennium bc, and evidence of a conflict that annihilated them. The most
obvious manifestation of warfare at Hasanlu is the complete destruction of
the late 9th century bc (period IVB) settlement by a conflagration, and the
interment within its ruins of the battles victims.
Ancient Near Eastern Warfare
The human actions that produced the archaeological remains at Hasanlu

can best be understood when viewed against the contemporary historical
background, specifically the 1st millennium Assyrian and Old Testament
systems of war and their treatment of the enemy (see below). The full force
and extent of the horrors of ancient Near Eastern warfare are presented to
us in gruesome and explicit details in the annals and records of the Assyrian
kings of the late 2nd and early 1st millennia bc. In the 13th century bc
Assyrian texts inaugurate what was to become a commonplace attribute of
1st millennium Assyrian records: the explicit description of battle slaughter
and the gory events that followed. King Shalmaneser I (ca. 12721244bc)
records that he blinded over 14,000 enemy prisoners; Assurbelkala in the 11th
century first mentions the flaying and impaling of prisoners; Tiglath-pileser
I (ca. 11141076 bc) mentions the deportation of captives.
* This article originally appeared as Warfare at Hasanlu in the Late 9th Century B.C.
Expedition, The University Museum Magazine of Archaeology and Anthropology, University of

Pennsylvania 31, nos. 23 (1989): 2436.
256 chapter seven
Figure 1. The Assryrian kings decorated their palaces with stone

reliefs showing their triumphs in battle. This scene, from the throne
room of King Assurnasirpal II (883859bc) at Nimrud, shows an
attack on a walled settlement defended by archers. Victims of the
battle fall from the battlements, and a woman m a tower holds her
head in sorrow (top center). Assyrian soldiers wearing pointed
helmets are undermining the walls (lower left) and attacking with an
armored battering ram (lower right). (Layard 1849; Pl. 19.)
In the 9th century bc the texts of Assurnasirpal II and Shalmaneser III

again record the mutilation of captives: the cutting off of their noses, ears,
limbs, impaling and blinding, and the immolation of male and female pris-
oners. Assurnasirpal II, like the Akkadians 1500 years earlier, mentions the
deliberate massacre of prisoners (see Schneider, this issue). Brutalities,
along with associated battle scenes, are also vividly and realistically depicted
on the bronze and stone wall and gate reliefs that decorated the 9th century
palaces of the Assyrian kings (Fig. 1). These reliefs served equally as histori-
cal records and as propaganda, inspiring both local and foreign visitors with
awe, and warning of the consequences to be suffered if they betrayed Assyr-
ian interests.
The Old Testament is another important source of information for war
and its tactics and brutalities in the early 1st millennium bc, especially for
the western states of the Near East-Israel, Edom, Syria. Aside from the usual
recording of warfare, we read also of the blinding or killing of prisoners of
war, and the slaughter of a captive citys population:
When you invest a city, you must offer it terms of peace . But if it will not
make peace with you, but wages war with you, you are to besiege it, and when
the Lord your God delivers it up to you, you must put every male in it to the
sword; but the women and children and live stock and everything that is in
the city, that is, all its spoil, you may take as your booty . [I]n the cities of the
warfare at hasanlu in the late 9th century bc 257
peoples here, which the Lord your God is giving you as a heritage, you must
not spare a living soul; but you must be sure to exterminate them, Hittites,
Amorites, Canaanites, Perizzites, Hivvites, and Jebusites, as the Lord your God
commanded you, so that they may not teach you to imitate all the abominable
practices that they have carried on for their gods, and so sin against the Lord
your God. (Deuteronomy 20:1018)
And ten thousand others did the Judeans carry away alive, and they brought
them to the top of a crag, and cast them down from the top of the crag, so that
all of them were dashed to pieces. But the men of the band whom Amaziah
had sent back without allowing them to go with him to battle fell upon the
cities of Judah from Samaria to Bethhoron, and slew of them three thousand
and took a large amount of spoil. (2nd Chronicles 25:1213)
(See also Josh 6:17, 24; Judges 1:6, 17:21; ISam 11:2, 30:17; II Kings 25:7.)
The Warriors of Hasanlu
The inhabitants of Hasanlu apparently did not differ greatly from their
neighbors in the emphasis placed on military activities. Lacking written
documents from Hasanlu itself, we do not know whether its foreign policy
involved a defensive strategy or an offensive one, or both. Artifacts found at
the site, however, certainly indicate that the inhabitants were prepared for
war. Weapons were recovered in all areas of the Citadel, inside and outside
buildings. In some cases, their number and disposition indicate that they
were found as they had been storedand never put to use in the final battle;
other isolated weapons were apparently abandoned or lost in the heat of the
battle.
Judging from the quantity excavated, the most common weapon em-
ployed at Hasanlu was the socketed spear, some made of bronze but the
majority of iron (Fig. 2); hundreds were found in the debris of the upper
floor of Burned Building II, where they apparently had been stored. Other
weapons include:
maceheads (Fig. 4)
48 stone
23 bronze
2 iron
3 iron/bronze
swords (Fig. 5)
2 bronze
28 iron
9 iron/bronze
1 with a gold cloisonn hilt
258 chapter seven
daggers (Fig. 5)
21 iron
1 bronze
3 iron/bronze
many blade fragments
bronze axes (Fig. 6)
iron pikes
Evidence for the use of the bow is provided by arrowheads (Fig. 7), some-
times found massed in a quiver. There were at least three bronze quivers, as
well as a unique iron quiver with bronze trim (Fig. 8).
Given this arsenal of weapons, it is not surprising that the Hasanlu finds
included protective metal armor. Among the bronze helmets were 2 crested
(Fig. 9) and 3 pointed (Fig. 11) examples, as well as 3 detached crests, their
(leather?) helmets having disintegrated; the crested helmets had separately
added ear guards. Three bronze shields, a pair of bronze shoulder guards and
a number of bronze and iron armor plates were also recovered. Even horses
sometimes wore armor on their heads (see de Schauensee, this issue).
Both military techniques and attitudes toward warfare can also be recon-
structed from representational art found at the site. In typical Near Eastern
fashion, the ruling elite of Hasanlu commissioned the illustration of bat-
tle scenes, which were displayed or stored in important buildings. These
scenes appear primarily on 72 small and fragmented ivory carvings that
were recovered (along with fragments bearing other motifs) from the second
story collapse of Burned Building II, identified as probably a religious struc-
ture or temple. (By contrast, only two battle scene fragments came from
Burned Building I, identified as probably the rulers residence.) Battle scenes
were also depicted on a silver and electrum beaker from Burned Building II,
and on five small rectangular bronze or iron plaques from Burned Building
I.
The ivory plaques had originally been attached with pegs to wooden
objects. Because the plaques had broken and scattered when they fell with
the collapse, the kinds of objects they decorated remain unknown, as does
the original grouping of the plaques and the way in which they were juxta-
posed. Lacking information on the relationship between plaques, we cannot
say to what extent they represent parts of a story or narrative. Both because
of their fragmentary nature and because none of the ivories has an inscrip-
tion (no local writing occurs at Hasanlu), we are also unable to identify or
explain a number of specific aspects of these scenes; for example, what is the
ethnic or political identity of the fighting forces, and what are the historical
event(s) represented?
Figure 2a, b. Spears made of iron or bronze with sockets that fitted over a
wooden shaft were the most common weapons found in the ruins
of Hasanlu. (a) The iron weapons were often in fragments, but
a complete example was found in Burned Building VI. (b) This
bimetallic spearpoint has a bronze blade riveted to an iron socket and
mid-rib. This weak construction and the rarity of such weapons
suggest they may have had a ceremonial rather than a utilitarian
function (see Pigott, this issue). (a: HAS 74249, L. 34.5cm, Muse Iran
Bastan, Tehran; b: UM 6531196, L. 57.0cm.
Photos courtesy of the Hasanlu Project.)
260 chapter seven
Figure 3. This reconstruction of a Hasanlu warrior with spear, small

shield, and feathered headgear is based on carved ivory plaque
fragments. The laced boots may not have been worn by soldiers,
but follow the form of a ceramic model. (Drawing by Ruth Stern.)
Figure 4ae. Maceheads from Hasanlu were made in a variety of forms

and materials, including stone, copper or bronze, and (rarely)
iron. Head wounds found on victims of the battle (Fig. 21)
testify to the use of such weapons to dispatch the wounded
and perhaps prisoners of war. (a: UM 615103, L. 7.9cm; b:
HAS 62339, L. 7.5cm; c: Metropolitan Museum of Art 61.100.17,
L. 8.5cm; d: Metropolitan Museum of Art 60.20.31, L. 8.5cm; e:
UM 595498, L. 6.0cm. Drawings courtesy of the Hasanlu Project.)
262 chapter seven
Figure 5ad. Daggers and swords were made of bronze or iron, and often
had decorated hilts, inlaid with ivory or wood. (a: Metropolitan
Museum of Art 63.109.4, L. 10.3cm; b: UM 5954129, L. 29.5cm; c:
HAS 59880, L. of hilt 23cm, Muse Iran Bastan, Tehran; d:
UM 615112, L. 41.0cm. Drawings courtesy of the Hasanlu Project.)
Figure 6a, b. Bronze axes such as these from Hasanlu were a traditional
weapon in Mesopotamia, depicted by the Sumerians as early as the 3rd
millennium bc. (a: HAS 621085, L. 18.0cm; b: HAS 58194, L. 13.5cm. Both
in Muse Iran Bastan, Tehran. Drawings courtesy of the Hasanlu Project.)
Figure 7ad. Arrowheads of iron and bronze were found throughout the
burned buildings. They vary in size and shape, so that a complete
study of their distribution within the site (including weapons in
storerooms) may allow us to distinguish between the arrows of
defenders and attackers. (a: HAS 74269, L, 9.9cm; b: HAS 74278,
L. 6.2cm; c: HAS 74290, L. 9.5cm; d: HAS 74276, L, 11.0cm. All from
Muse Iran Bastan, Tehran. Photos courtesy of the Hasanlu Project.)
264 chapter seven
Figure 8. Empty quivers were apparently discarded during the course of

the battle, and this unique example lay on the paving of the Lower Court.
Made of iron with bronze bosses and tacks, it presumably had a leather
lining. The iron sheeting was decorated with hammered (repouss)
fluting and figures of people and animals (see Pigott, this issue).
L. 58.3cm. (UM 7123324; photo courtesy of the Hasanlu Project.)
Figure 9a, b. Bronze helmets with crests were sometimes elaborately

decorated with fine incised designs (b). Holes along the edges
of the helmet, as well as on the edges of metal ear flaps, indicate
the presence of cloth or leather linings. Plain leather(?) helmets
must also have been worn, documented by metal crests and flaps
(a). (a: Metropolitan Museum of Art 61.100.39, L. of ear flap
14.0cm; b: HAS 60528, Ht. of helmet 31.0cm, Muse Iran
Bastan, Tehran. Drawings courtesy of the Hasanlu Project.)
266 chapter seven
Figure 10. The archer in this reconstruction wears a costume like

that depicted on the silver cup from H asanlu (see Marcus, this
issue). His iron arrow and bronze quiver reproduce artifacts
recovered from the burned settlement. (Drawing by Ruth Stern.)
Figure 11 a, b. Pointed bronze helmets of the type depicted on Assyrian

troops in 9th-century reliefs (Fig. 1) may also have been worn by the
soldiers of Hasanlu. Several archaeological examples were found (a), and
the central figure on a decorated breastplate from Hasanlu wears
similar headgear (b). (a: UM 615352, Ht. 26.0cm; b: HAS 74241,
Muse Iran Bastan, Tehran. Photos courtesy of the Hasanlu Project.)
Figure 12a. Scenes of battle are found in Near Eastern art as early as the
4th millenium bc. An impression made by a cylinder seal from the site of
Chogha Mish in southwestern Iran shows human figures on the walls of a
fortress, hands raised in distress or surrender. Two larger figures (left)
represent the conquerors, one of whom holds a small seated(?) captive by
the hair. (Drawing by Jon Snyder after Delougaz and Kantor 1972: Pl. 10:d.)
268 chapter seven
Figure 12b. A clay sealing from the site of Susa in southwestern Iran shows
a citadel, its walls ornamented with horns. Three nude figures
represent the defeated forces, falling or fleeing before a fully clothed
archer. One of the small figures appears to have his hands bound
behind his back. (Drawing by Jon Snyder after Amiet 1966: Fig. 11.)
Figure 12c. A sealing from Chogha Mish shows a triumphant ruler

returning home with his booty. He holds a mace in one hand, while in the
other he holds a rope around the neck of a seated prisoner of war.
(Drawing by Jon Snyder after Delougaz and Kantor 1972: Pl. 10:c.)
Figure 13. During the 3rd millennium, Mesopotamian rulers recorded their
feats on sculpeted stone monuments. The Stele of the Vultures shows
the army of the city of Girsu trampling over the defeated army of Umma.
(The Louvre: pho courtesy of the Metropolitan Museum of Art.)
270 chapter seven
Figure 14. Some of the people who died in the columned hall of Burned
Building II were richly adorned with jewelry. The skeleton shown here
belonged to a young woman in her late teens. She had a necklace of
amber, carnelian, frit, and shell beads, and other strung beads lay below
her pelvis. On her upper body were three lion pins bearing cloth
impressions, one of them visible here (see Pigott, this issue). At her knee
was an iron dagger with a bronze hilt; traces of a scabbard adhered to
the blade. (HAS skel. 263. Photo courtesy of the Hasanlu Project.)
Figure 15a, b. Small ivory plaques from Hasanlu depict mounted

soldiers with spears and chariots trampling over the bodies of the
enemy. Following Mesopotamian tradition, the defeated soldiers
are nude. The Metropolitan Museum of Art, Rogers Fund, 1965
(65.163.9, .19, .21) Image The Metropolitan Museum of Art.
272 chapter seven
Figure 16. Vigorous battle scenes depicting foot soldiers fighting with
spears and small shields are found on the Hasanlu ivory plaques.
The man on the left wears a crested helmet (see Fig. 9). Ht. upper
frag. 3.4cm. The Metropolitan Museum of Art, Rogers Fund, 1965
(65.163.8, .18) Image The Metropolitan Museum of Art.
Figure 17. This ivory fragment shows a bare-headed soldier

attacking chariot horses with a spear. Ht. 4.0cm. (HAS 70409,
Muse Iran Bastan, Tehran. Photo courtesy of the Hasanlu Project.)
Figure 18. Fragment of a siege scene showing an attacker attempting to

scale a wall or platform with a ladder. The feet of the defenders,
one of whom carries a shield, are visible at the top. Between the
fortress and ladder a horses head is visible. Ht. 3.7cm. (HAS 60950,
Muse Iran Bastan, Tehran. Photo courtesy of the Hasanlu Project.)
274 chapter seven
Figure 19. Women are rarely depicted in the art of 9th century Hasanlu, in
which scenes of warfare and ceremony predominate. This ivory
plaque fragment shows a figure with braided hair; both the hairstyle
and the position of the hands suggest this is a woman, in
despair as she watches the enemys attack. An arrow juts out of
the tower behind her. Ht. 3.0cm. (Metropolitan Museum of Art
65.163.24. Photo courtesy of the Metropolitan Museum of Art.)
Through analogy with other Near Eastern representations, we may

assume that the victorious, energetic attackers are the locals (probably but
not necessarily the Hasanlu army), and that those they oppose and those
trampled under chariot and cavalry horses are the enemy (Fig. 15). But aside
from this basic information we are able to study the scenes only formally,
and in isolation. Collectively, the battle scenes depicted on the ivories follow
the conventional patterns known to us from Assyrian texts and representa-
tions, with only minor variations in details. The military forces represented
are infantry, chariotry, and cavalry for the locals, and infantry and perhaps
chariotry for their antagonists.
The Hasanlu infantry wore belted kilts that may have been protected
by armor plates, feathered (or crested) helmets, sometimes with ear guards,
and they fought barefoot or with sandals. Their weapon was the spear,
perhaps also the sword, and they carried small shields (Figs. 3, 16). No
axes, daggers, or maces are evident in the plaque scenes, although as noted
above, these weapons were recovered at Hasanlu. The diameters of the 3
shields excavated at Hasanlu are respectively 33, 37, and 43 cm; the first
two, at least, could be the same size as those represented on the ivories.
That the belts worn may have been metal is suggested by the many bronze
examples excavated at Hasanlu. It is almost impossible to determine how
the enemy infantry was outfitted, but one certain representation shows an
unhelmeted figure wielding a spear against local chariotry (Fig. 17); another
fragment shows barefooted antagonists fighting each other (shown with
another fragment in Fig. 16).
On the ivories, the Hasanlu cavalry employs only the spear; this contrasts
with the way in which Assyrian reliefs display their cavalry, shown wielding
the spear, the bow, or both. The Hasanlu riders wore boots, and seem to
have ridden bareback since no saddles are depicted (Fig. 15b). Chariots were
driven by two horses and staffed by two men, the rider and an archer
paralleling Assyrian and North Syrian custom for this period. Chariot wheels
had six spokes (Fig. 15a), although one fragment shows a four-spoked wheel,
perhaps an indication that this is an enemy chariot. Unlike Assyrian custom
of the 9th century, but similar to contemporary North Syrian cavalry, there
were no outrider horses used for the cavalry or chariots, nor are blinkers,
bits, or horse armor depicted.
Two, possibly three, ivory fragments depict the siege of a city, a common
motif on Assyrian reliefs. On one fragment a ladder placed against a plat-
form is being mounted by an attacker who confronts a defender (Fig. 18).
Another poignantly depicts a female holding her head in grief, next to a city
turret just struck by an arrow (Fig. 19; see also Fig. 1). On a bronze plaque, four
276 chapter seven
archers shoot from platforms (like those of Fig. 18) seemingly set above forti-
fication walls, probably representing a city under siege. Other metal plaques
depict archers, cavalry, and chariots, but not in scenes where an enemy is
present. Only in one instance, on a silver and electrum beaker found at
Hasanlu (see Marcus, Fig. 1), is a prisoner of war apparently represented, for
here an armed soldier clasps the wrist of an unarmed figure.
As fragmented as they are, the ivory carvings enable us to present some
observations about battle arrangements. Cavalry and chariots are shown
fighting in action together side by sidedepicted on the carvings by over-
lapping figures (Fig. 15a, lower right; note the hooves of a cavalry horse
directly in front of the chariot horses). Likewise, infantry are shown fight-
ing alongside chariot and cavalry horses (Fig. 15a, upper right; the hand of
an infantryman wielding a spear may be seen in front of the two chariot
horses heads). Several fragments depict soldiers attacking cavalry and char-
iot horses (Fig. 17), and others show infantry fighting hand to hand (Fig. 16).
The cavalry, albeit without stirrups for support, seem to have functioned as
lancers (Fig. 15b).
The distinct impression that one derives is that chariots, cavalry, and
infantry seem to have participated in battle together without functional
division- but it should be noted that the depictions may be merely an artistic
convention and not necessarily a literal recording of battle maneuvering. It
is thought by military historians that before the late 1st millennium bc (the
Achaemenid period) chariots functioned primarily to transport troops to
battle and to harass the enemy with arrows (see Xenophon Cyropaedia VI,i:
2730). The chariots were not to get too close to their infantry, because of
the vulnerability of the horses to enemy weapons. The Hasanlu scenes do
not seem to conform to this concept, although in fact, we may be viewing a
representation of a mopping-up activity against a fleeing, broken enemy.
Light war chariots had been functioning in battle throughout the Near
East for about 1000 years before the destruction of Hasanlu, but the use of a
cavalry corps is not attested before the time of the Assyrian king Assurnasir-
pal II in the early 9th century bc. During his reign and that of his succes-
sor Shalmaneser III, cavalry is often mentioned in battle descriptions and
depicted in art. In addition to its use in Assyria, a cavalry corps is recorded
as forming units in several other armies, including those of Urartu, Baby-
lon, Adini, Israel, and Syria. In the 9th century bc, then, both cavalry and
chariotry were normal constituents of a Near Eastern army. A large num-
ber of horse bits, protective breastplates, and harness equipment, as well as
the skeletons of horses have been excavated at Hasanlu. A fragmented mass
of wood together with a long pole from Burned Building IVV may be the
remains of a chariot; and some bronze and carved stone objects that may be
yoke saddle pommels for chariots were recovered (see de Schauensee, this
issue). Thus, the archaeological remains support the pictorial evidence that
chariots and cavalry functioned as elements of the Hasanlu army.
It need not be assumed that the representations of battle scenes on the
ivories were conceived by the local authority solely to emulate Assyrian
propaganda techniques projecting power and royal might. And it is prob-
able that the military forces of Hasanlu itself are represented fighting and
defeating an historical enemy. If so, the depictions of a cavalry and chariot
corps signify that there existed at Hasanlu an elite class that had the leisure
and skills to practice and perform the necessary complex maneuvering tac-
tics. Whether this class itself maintained and supplied the equipment and
horses or whether this was a state function is unknown. Equally unknown is
whether there was a conscript or a standing army, the latter being a standard
feature of the Assyrians by this time.
A Brief History of Warfare in Mesopotamia (Excursus)
Evidence for warfare in Mesopotamia and adjacent regions begins in pre-

historic times, very soon after people began building permanent settle-
ments. Among the earliest fortified towns and villages are the early Neolithic
site at Jericho dating to the 8th millennium bc; late Neolithic-Chalcolithic
Hacilar in Anatolia, 6th millennium bc; Tell-es Sawwan in Mesopotamia,
ca. 6000 bc; Ugarit in Syria, and Mersin in Cilicia, ca. 5000 bc. Archaeolog-
ical evidence also documents settlements that were destroyed by fire and
abandoned, commonplace events that in some cases surely reflect hos-
tile human activity. Fortified sites with destruction levels are recorded by
archaeological research throughout the greater Near East for many millen-
nia.
Ancient pictorial representations also furnish information about war.
Seals and seal impression dated to the Uruk period (late 4th millennium bc)
in southern Mesopotamia and southwestern Iran depict battles and their
consequences. Seals from the sites of Susa and Chogha Mish in Iran show
scenes of battle and the taking of a walled citadel (Fig. 12a). Contemporary
seals from the city of Uruk in Mesopotamia and from Susa depict men in
undignified positions and with their arms tied behind their backs, scenes
that may represent prisoners of war rather than civil criminals (Fig. 12b).
A sealing from Chogha Mish shows a large seated figure in a boat, plau-
sibly interpreted as a ruler returning home with his booty. In one hand
278 chapter seven
he holds a mace, while the other grasps a rope attached to a pair of seated
captives (Fig. 12c).
The earliest historical texts in cuneiform, written during the Early Dynas-
tic Period in Sumer (ca. 2900 to 2350 bc), contain records of war between
the independent city states of southern Mesopotamia and exaltation in the
slaughter of the enemy. Contemporary sculptures depict warfare in vivid
detail, and elucidate the cuneiform texts. One of the best-known monu-
ments of the Early Dynastic period, the Stele of the Vultures (Fig. 13), was
carved for King Eannatum of Girsu. On the stele are scenes showing dis-
ciplined uniformed troops from Girsu carrying spears and battle axes as
they trample over the nude bodies of the soldiers of Umma; nearby are
nude Umma bodies piled in a heap, while others are being buried under
a mound of earth. A mosaic panel from the Royal Cemetery at Ur (the Ur
Standard) shows four-wheel warwagons staffed by lancers that override the
nude bodies of an enemy, and a procession of prisoners, their arms pin-
ioned behind their backs. From this same period, shell plaques from Mari on
the Euphrates River depict both soldiers bearing axes and bound prisoners.
The portrayal of enemy dead in the form of nude bodies being overridden
by victorious forcesusually chariots or cavalryand of bound prisoners
became standard victory motifs that continued to be represented in Near
Eastern art the the 1st millenium bc.
Beginning in the period of Semitic Akkadian control of Mesopotamia
(23342154bc) prisoners of war are mentioned for the first time in texts, and
Akkadian reliefs dramatically depict them with their hands or arms held
by rope behind their backs, and sometimes with their necks held in stocks.
Also during this periods texts mentions the slaughter of prisoners for the
first time; such an act is possibly portrayed on an Akkadian stele from Girsu
that shows soldiers killing unarmed nude men. Deportation of the captured
inhabitants of cities is first mentioned in the later texts of the Sumerian 3rd
Dynasty of Ur period (ca. 21122004bc).
The Battle of Hasanlu and Its Victims
Approximately 246 skeletons of men, women, children, and infants were

recovered, who perished either as a result of the fire, or as targeted vic-
tims of violence. A number of the unfortunates (about 157) were found in
five of the burned buildings where they were caught when the structures
collapsed. For example, in the Great Hall of Burned Building II about 50
victims were uncovered lying clustered near the main (northern) doorway,
Figure 20. Victims of the battle at Hasanlu were often found trapped
within buildings that had burned and collapsed. This group lay
at the northern end of the columned hall in Burned Building II,
near the outside door. Other items found on the floor included
charred beams (upper left) and a red deer skull with antlers (in
front of the workman). (Photo courtesy of the Hasanlu Project.)
crushed beneath fallen walls and roof (Fig. 20). The victims included men,
women, and many children. Some individuals were armed, while many of
the females and children were wearing jewelry, including relatively heavy
lionheaded pins (Fig. 14). Another group of about 89 people were found
where they fell in open areas, victims of slaughter. The cause of death is
graphically documented by head wounds, or by disarticulated limbs; in the
latter cases, the bodies could have been mutilated by animals and vultures,
but the head wounds patently tell another story. In each category, those who
perished from the collapse of the buildings, and those from slaughter, were
men, women, children, and infants.
Three clusters of skeletons revealed chilling episodes of death at Hasanlu.
In the first two cases, random slaughter of fleeing troops and inhabitants
seems to have occurred. South of Burned Building XI the skeletons of 11
280 chapter seven
Figure 21. A large group of people were slaughtered outside the entrance
to the settlement, to the south of Burned Building XI. Several in the group
had head wounds, visible on the skulls as bone damage. Skeleton 392, a
male of about 20, was killed by a mace that left a rounded depression on
the back of the skull (right). A healed lenticular (sword) wound is also
visible in the center of the photo; his age and the earlier wound suggest
that he may have been a soldier. (Photo courtesy of the Hasanlu Project.)
adults, 3 teenagers, 1 child, and 1 infant were uncovered close to one another;
none was armed. Six of the adults had head wounds, some multiple (Fig. 21).
The others were probably killed by wounds to other areas of the body, which
would not necessarily leave traces on the skeleton, a situation that no doubt
applies to other skeletons uncovered. And to the east of the Upper Court
Gate were uncovered the clustered skeletons of 6 adults, 2 teenagers, and 1
infant; one of the adults had a head wound.
The third group of skeletons almost certainly represents slaughtered pris-
oners. In Room 2 of Burned Building IV the skeletons of 27 people were
discovered16 adults, the rest children and infants. Found overlapping or
close together in a mass, the skeletons were lying in disarray over the burned
debris from the collapsed roof, and were covered by a layer of brick debris.
Four of the victims had head wounds, unambiguous evidence that they
had been killed by weapons; the form of bone damage indicates that the
weapon(s) used, perhaps to dispatch an already wounded and helpless per-
son, was a mace.
The fact that the victims were lying on burned debris indicates that they
were killed after the initial collapse of the building, but while the walls were
still standing. This evidence suggests that the victims seem to have been
purposely brought to Room 2 to be executed shortly after the fire had died
out and the destruction had been accomplished. At some time after their
death the walls collapsed and sealed them in the rubble.
Whether the killing of prisoners (if indeed such was the case here), as
opposed to the random slaughter of fleeing inhabitants, was an isolated
event at Hasanlu is unclear. Nothing is known of the dynamics that pre-
ceded the invasion of the Citadel by the enemy: whether the inhabitants
surrendered only to be killed nevertheless, or whether they did not surren-
der, and being defeated suffered the consequences. In this context we may
wonder if the massacre in Burned Building IV was a punitive political action
or a religious one, such as sacrifice?
Likewise we wonder about the presence or absence of the Hasanlu armed
forces: were they defeated in the neighboring plains, or were they absent on
a campaign? Wherever they were, they could not prevent the invasion, an
invasion that was accomplished by a strict following of the Assyrian method
of warfare.
The fire that destroyed the site at Hasanlu either was deliberately set or
resulted from an accident, but it must have started soon after the enemy
entered the city. Whatever the cause, the fire spread through the wood
and brick buildings of the Citadel quite quickly, apparently preventing the
enemy from acquiring much, if any, booty. There is no obvious evidence
within the debris that looting or post-destruction digging for treasure
occurred; rather the opposite is suggested. For example, the well-known
gold bowl (actually a beaker; see Winter, this issue) was found in the arms
of, apparently, a local inhabitant, who in the process of attempting to save
it died in the collapse of Burned Building I-West. And the vast quantity of
material recovered within the debris of all the buildings, some of it sump-
tuous (artifacts of silver, gold, ivory, Egyptian Blue), suggests that the Citys
contents remained essentially unplundered.
282 chapter seven
Who Burned Hasanlu?
To encounter the material remains of the destruction of the buildings and

inhabitants of Hasanlu is to confront the actualities of the written descrip-
tions and illustrated reliefs of the Assyrians: burning, killing, massacre. Not
a few archaeologists who excavated at the site were emotionally affected by
the carnage and the human suffering that had taken place. Yet, there is no
historical evidence that it was the Assyrians who destroyed Hasanlu. In fact,
both locally derived chronology and archaeological and textual evidence of
Urartian penetration into the neighboring Ushnu valley directly to the west
(at the mountaintop site of Qalatgah) suggest that it was an Urartian army in
the last decade of the 9th century bc that probably destroyed Hasanlu. This
information soberly expands our perceptions, geographically and culturally,
about the extent and nature of the horrors of warfare in the first millennium
bc.
Urartian texts, while not so graphically descriptive as the Assyrian
sources, do mention some Urartian military tactics, including the taking of
prisoners in battle; and at least one text from the 8th century bc refers to the
deportation of captured troops. Perhaps, then, we may assume that some
prisonersif not bootywere taken by the Urartians at Hasanlu. Urartian
texts also reveal that the king himself led his forces in battle. It was King Ish-
puina and his son Menua, the builders of Qalatgah in the late 9th century bc
during their joint reign, who may have been responsible for the attack on
Hasanlu. If so, then we may assume that both these kings were present dur-
ing the sites destruction. And if this interpretation is correct, the destruction
and slaughter at Hasanlu IV furnishes grisly information about the customs
associated with a Urartian military campaign. Ironically, excavations at the
Urartian site of Karmir Blur in Soviet Armenia reveal that when that city was
destroyed in the last half of the 7th century bc, it too experienced violent
burning and slaughter.
Not knowing either the ancient name of the city, or its political affiliation,
we are frustratingly unable to relate it historically to one of the many cities
or states situated in Iran that are mentioned in Assyrian texts. Yet, as noted
above, the historical and archaeological evidence suggest that the enemy
that destroyed Hasanlu was Urartu. If so, we may ask if there is any evi-
dence for the presence of Urartian troops within the ruins of Hasanlu. The
answer is no, except for possible but tenuous indications or clues. One pos-
sible clue is the presence of five complete or fragmentary crested helmets
(Fig. 9), which have been called U rartian by R.D. Barnett and J. Borchhardt.
Crested helmets, without ear guards, were worn by North Syrian troops, and
also by those Urartians represented on the bronze gates from Balawat com-
missioned by Shalmaneser III (858824 bc).
The Urartian helmets on the Balawat gates, however, are depicted with
short, pointed ear guards made with the helmets, not added on as in the
Hasanlu examples, and thus it cannot be argued with force that the latter
belonged to the Urartian invaders. And if the ivory pictorial evidence indi-
cates that the Hasanlu infantry wore feathered/crested helmets, they are not
the same as the excavated examples. Who, then, wore the crested helmets
recovered at Hasanlu? Equally puzzling is the presence of the pointed hel-
mets (Fig. 11), which are very like the Assyrian standard form, and which after
the mid-9th century were represented in art as also worn by Urartians. Did
the Urartian troops wear both crested (for officers?) and pointed helmets
while at Hasanlu? Or, did the local forces themselves employ a variety of
helmet forms? More research and thought are needed to resolve this prob-
lem.
Another possible clue depends on the cultural attribution of a bronze
mace head (Fig. 4c) with star or rosette faces. Two very similar examples,
albeit found in a later context, derive from Altin Tepe, an Urartian site in
northeastern Turkey. Were the Altin Tepe examples booty from the Hasanlu
campaign kept as heirlooms, or were they locally made in Urartu under
northwestern Iranian influence; or is the Hasanlu example a weapon lost
there by an Urartian soldier? We do not know. Nor in fact do we know how
many (if any) of the other artifacts excavated at Hasanlu actually may have
been left by the enemy forces and are thus not to be documented as local
products. Here too, more research, not speculation, is required to resolve
this tantalizing issue.
From all the evidence made available by archaeologythe destruction,
the artifacts, the pictorial representationsit is attested that warfare was
not a casual or incidental activity for the people of Hasanlu IV. Nevertheless,
archaeology has also revealed that there was time, energy, and talent for
architects and workmen to construct monumental buildings, and for highly
skilled craftsmen to manufacture a large variety of objects, both luxury items
and objects for daily use. At Hasanlu there was a time for war and a time for
peace, but war was the ultimate event in its history.
284 chapter seven
Bibliography
Arniet, Pierre
1966 Elam. Paris: Centre National de la Recherche Scientifique.
Brentjes, Burchard
1986 Kriegwesen im alten Orient. Das Altertum 32:134142.
Delougaz, P.P., and Helene J. Kantor

1972 New Evidence for the Prehistoric and Protoliterate Culture Development
of Khuzestan. In Vth International Congress of Iranian Art and Archaeology:
Tehran, Isfahan, Shiraz; Vol. 1, pp. 1433. Tehran: Ministry of Culture and
Arts.
Dyson, R.H., and Muscarella, Oscar White

1989 Constructing the Chronology and Historical Implications of Hasanlu IV.
Iran 27:127.
Gelb, I.J.
1973 Prisoners of War in Early Mesopotamia. Journal of Near Eastern Studies
32(12):7098.
Layard, Austen Henry

1849 The Monuments of Ninevah. Vol. 1. London: John Murray.
Muscarella, Oscar White

1980 The Catalogue of Ivories from Hasanlu, Iran. Philadelphia: The University
Museum.
Rolle, R.
1977 Urartu und die Reiternomaden. Saeculum 28(3):291339.
Schulman, A.R
19791980 Chariots, Chariotry, and the Hyksos. Journal of the Society for the
Study of Egyptian Antiquities 10:105153.
von Soden, Wolfram

1963 Die Assyrer und der Krieg. Iraq 25(2): 131144.
Yadin, Yigael
1963 The Art of War in Biblical Lands. 2 vols, New York: McGraw Hill.
chapter eight
THE HASANLU LION PINS AGAIN*
In August of 2001 Charles Burney returned to the city of Van/Tushpa to

revisit the region in which he made major archaeological contributions. The
occasion was to attend the Fifth Anatolian Iron Ages Symposium, hosted
by Prof. Dr Altan ilingiroglu. Here a paper was presented by Michelle
Marcus and Karen Rubinson (delivered by the latter), which reported on
Iranian, Anatolian and Caucasian jewellery. One topic discussed was the
excavation background and cultural message of bronze lion pins: a solid
bronze recumbent lion cast on to an iron tang pin that fastened clothing
with the help of an attached chain. The lions differ in size, weight, body
decoration and modelling (Fig. 1), and to date have been recovered only
at Hasanlu, Period IV, in Northwestern Iran.1 Marcus-Rubinson cited only
those examples recovered in association with skeletons excavated in the
pillared hall of Burned Building II (BB II), destroyed in a general confla-
gration of ca. 800 bc. The specific issues and conclusions they presented
were:
a) that many of the skeletons discovered crushed in the collapsed debris
were recovered with lion pins on their bodies.
b) that a crushed skeleton (skeleton 263, shown in the lecture in the
process of being excavated),2 recovered with three lion pins touching
the upper part of the body, was that of a female.
* This chapter originally appeared as The Hasanlu Lion Pins Again, in A View from the
Highlands: Archaeological Studies in Honour of Charles Burney, ed. A. Sagona (Leuven: Peeters,
2004) 693710.
1 Muscarella 1988, pp. 4245 for discussion; Marcus 1993. Some years ago in a Van jew-
ellers shop I saw a typical Hasanlu lion pin that the shopkeeper told me came from a Turkish
site north of Van. This said to be from site X was an attempteither by the original seller
or the shopkeeper, or bothto cover the fact that the pin (and hundreds of other ancient
and modern objects sold in Turkey) was smuggled from Iran. For other plundered strays see
below and note 13.
2 Previously published in Muscarella 1989, p. 32, fig. 14. The upper figure in this photo-
graph is the present author; the other is Noruz from the village of Hasanlu, a superb excavator.
This photograph was taken shortly before Marcus 1993, p. 160, fig. 2 (Fig. 2 here): in the former
photo the lion pin near the elbow was only just emerging.
286 chapter eight
c) and, to document how lion pins were worn, a drawing was shown
depicting a female, arms akimbo, who wears bracelets, rings, several
necklaces, and two lion pins athwart one shoulder and one on the
other.3
In the discussion period, I confronted these assertions. Here I expand what
I stated in Van, but I do so by responding to a written paper, for the verbal
claims made in Van had appeared earlier and in more detail in Marcus 1993
(but I cannot avoid responding to brief discussions of lion pins in three other
articles published by Marcus, in 1994, 1995 and 1996). I arrive at different
conclusions and interpretations from Marcus. I argue here that a review of
the excavated evidence and data indicates that a number of statements and
cultural conclusions made in the various published papers reflect neither an
accurate archaeological history nor cultural explanation for the presence of
lion pins at Hasanlu.
It was at Hasanlu where I first met Charles and it is appropriate as well as
personally satisfying to present to him a work that addresses a special group
of artifacts from that extraordinary site.
Marcus argument can be summarised as follows: 60 adorned individu-
als were recovered within the ruins of BB II, of which, about half were
wearing lion pins, or associated with lion pins; that 69 of the approxi-
mately 100 examples recovered from the site at large derived from BB II;
and that most [of the 69] were found on skeletons [italics added].4
Concerning the number worn, she states that children wore one lion pin,
claiming five children were recovered wearing lion pins, and older children
and adults wore two or three pins.5 With regard to their positioning on the
bodyJudging from the distribution of the lion pins on or near the upper
body [she cites her fig. 2, which is Fig. 2 in this paper],6 they were probably
originally pinned at the chest and shoulders and Marcus again records
their placement on the upper body 7
3 Previously published in Marcus 1993, p. 171, fig. 14.

4 Marcus 1993, pp. 159, 170, 163; 1995, pp. 2499, 2592. But we are told in 1995, p. 2502 that
in BB II pins were found on at least twenty-four skeletons.
5 Marcus 1993, p. 164; but in 1995, p. 2497, she makes a generalisation that individuals
wore three or more and on p. 2502 that they were worn, in groups of three
6 Her fig. 2 had been previously published in Dyson 1964, p. 374, fig. 12, and Muscarella
1988, p. 44, fig. 5, which is skeleton 263. In Dysons fig. 12 the caption mentions the two visible
lion pins; in his fig. 9 the caption mistakenly states that the drawing is of the pin in fig. 12.
The drawing is not from skl 263, but an isolated example from BB II (60954): see Marcus
1993, p. 164, fig. 6. For the lion pins from skl 263, see Marcus 1993, p. 167, fig. 9.
7 Marcus 1993, pp. 164, 172; 1995, p. 2502: note the plural shoulders, and see infra.
the hasanlu lion pins again 287
The archaeological information obtained from my examination of the

Hasanlu field records is sometimes different from these claims, in some
instances significantly.8 There were a total of 62 skeletons recovered in BB II,
of which at least 18 (skeletons 80, 120, 124, 131, 132, 151, 153, 159, 257259, 261,
323, 437, 439,513515) were recorded in the field as unadorned, i.e., with no
adornments or artifacts on or near them. Of these unadorned skeletons, four
were listed as adultsone anthropologically sexed as a female (skeleton
261); and three as children (skeletons 120, 124, 513)one very young (skele-
ton 124).
In the Mortuary Furniture section on the field report sheets (where
material associated in some manner with the skeleton was recorded), 28
skeletons9 were registered with various types of jewellery such as necklaces,
bracelets, rings, but no lion pins. One of these skeletons was recorded as
having nothing present but a crested helmet (skeleton 126).10 Identified
among this group were nine adults, one anthropologically sexed as a male,
four young children (skeletons 121, 122, 123, 125), a child (skeleton 146), and
one infant (skeleton 148).11
Thus, there were not 60 adorned individuals in BB II; and 46 of the
62 excavated skeletons were recorded in the field reports as not adorned
with lion pins at the time they died, crushed by the collapse of BB II. This
calculation leaves a possible 16not 30excavated skeletons adorned with
lion pins, but as we shall see there is also a problem of specific numbers here
when one examines the evidence recorded in the field.
The number 69 for the quantity of lion pins recovered in BB II is cor-
rect (which corrects Muscarella12 based on wrong information). The total
number of lion pins excavated at the site at large is 103. Of these, 97 were
excavated by the Hasanlu Project, and one was brought by a local villager
to the excavators in 1956. A. Stein excavated one example in 1936 and in the
commercial excavations of 1947 and 1949, four examples were excavated
8 I thank Robert Dyson for giving me much of his time to provide me with photographs,
field record sheets recording all the BB II skeletons, and answers to my many questions
promptly. Shannon White of the University Museum of the University of Pennsylvania also
generously helped me on numerous issues.
9 Skeletons 81, 82, 121123, 125130, 133, 136, 137, 142, 143, 145149, 154, 155, 260, 262, 435,
436, 438.
10 Muscarella 1989, p. 29, fig. 9b.
11 For skeletons 123, 260, 262 in situ in BB II see ibid., fig. 20, lower right, lower left, top
right.
12 Muscarella 1988, p. 42.
288 chapter eight
from the outer town area; they are now in Teheran. [Unexcavated, purchased
examples include one example in the Louvre (acquired in 1958: AO 20472,
ex-collection Coiffard), two examples in Los Angeles, another in a Japanese
collection, one seen by me in Van (see note 1), and at least three that were
advertised in the bazaars.]13
In the mortuary furniture lists for 13 of the BB II skeletons (skeletons 119,
134, 135, 138141, 144, 150, 152, 156158), which include five adults and three
children (skeletons 119, 144, 156), are listed bracelets, necklaces, weapons,
etc., as well as lion pins. However, in not one instance did the excavator men-
tion precisely where, on or near (or how near) the skeleton, the objects, were
recovered (see also below). One may surmise that the recorded bracelets,
anklets, and rings were indeed found on their expected positions; for other
artifacts recorded, the situation is more difficult. In some cases it is clear that
artifacts recovered in the vicinity of a skeleton were recorded as belonging
to that particular skeletons personal adornment (however, specific loci on,
or measurements of distances from, a skeleton were not given). For exam-
ple, skeleton 158 has extant only its legs and part of its pelvis, with no body or
skull: where specifically was its furniture, listed as beads and one lion pin, in
fact recovered? In one case (skeleton 141) the excavator reported 9 lions in
the furniture inventory, surely an incorrect number with regard to on-body
placement. Although we remain ignorant whether any of these pins were
recovered actually on skeleton 141for no loci positions were recordedwe
can claim based on the limited empirical evidence at Hasanlu (see below)
that at least six were not recovered on the skeleton, but recovered some-
where in its vicinity. The same excavator accurately recorded that six lion
pins were recovered in the vicinity ofnot onsix skeletons (skeletons
128133), and suggested in this instance that they prob[ably] belong to
them. This seems to be a viable interpretation but we cannot speculate
how many of the six lion pins belonged with each of the six skeletons.
13 The sum 103 with details comes from R.H. Dyson, personal communication, and Mus-
carella 1988, pp. 43, 45 and note 2 for published records of the unexcavated examples; also
A. Hakemi and M. Rad, The Description and Results of the Scientific Excavations of Hasanlu,
Solduz, Teheran 1950, figs. 31, 32 for four excavated examples. I do not know whether an exam-
ple once in the Brummer collection later surfaced in one of the mentioned collections, The
Los Angeles County Art Museum for example. For the Louvre example see Ghirshman 1964,
p. 27, fig. 28. Most probably these unexcavated examples derive from Hasanlu. Hence to the
103 excavated number we may add 7 or 8 others to the corpus. However, the problem remains
whether other Iron II sites in the area (viz. Nagadeh, Dinkha, and Geoy) may also have used
lion pins, and thus it would be too absolute to claim that all the orphans must have derived
from Hasanlu.
These examples, plus the lack of specific loci for the other listed lion pins,
compromise our knowing how many of these 13 skeletons actually wore
lion pins, and how many were worn per skeleton (see also skeleton 255,
below).
In fact, only three skeletons (skeletons 255, 256, 263)14 were explicitly
recorded as having lion pins on or immediately contiguous to their physical
remains. Skeleton 255 (a young adult) had a lion pin, pin towards body,
resting on its left humerus, and two pins found adjacent to but not touching
skeleton. However, a photograph (Fig. 3) and a sketch made in the field show
the former, but nothing of the other two is visible or drawn near the body
(this could signify that they were not found in the immediate vicinity, or
were removed before the photograph was taken; probably the latter).15 The
second skeleton, a child (skeleton 256), was recovered close to skeleton 255;
recorded in its inventory is a lion pin touching its only preserved humerus
and another next to the body. The field sketch shows one on the humerus
and another just above and parallel to it, close to the upper chest; clearly
both pins belonged to skeleton 256. The third, a young adult (skeleton 263,
Fig. 3) had three lion pins on its body: one touching the chest, pin up,
one recovered touching the inner side of the right arm, pin up, and a third
recovered at neck/shoulder area.16 From these field descriptions, skeleton
255s in situ pin would have been attached to clothing on some part of the
upper body, probably close to or on the shoulder, or near the neck. Skeleton
256s two pins were fitted one above the other at the chest. Skeleton 263 had
them in three superimposed rows, from the shoulder/neck to the waist area.
The field-report sheet for another skeleton (skeleton 157) records a
bronze lion under leg, a locus that could suggest either that it belonged to
that skeleton or had earlier fallen from another fleeing individual. Notwith-
standing the field record, however, a photograph of the skeleton in situ
shows a lion pin on the floor ca. 1220cm. to the left of its left leg (Fig. 4).
14 I excavated these three skeletons in 1962but except for skeleton 263 have no memory
of the excavation details. Several individuals and I excavated the others mentioned in this
paper in 1960 and 1962.
15 Note that a garment pin was also recorded with skeleton 255 as near the arm but it too
does not appear in Fig. 4. The problem of recognizing discrete skeleton inventories was quite
difficult given the mass of skeletons and scattered body bones (not to mention intruding
emotional feelings over the chaos we were uncovering and witnessing).
16 Muscarella 1989, p. 32, fig. 14.
290 chapter eight
So, what may one conclude about the number of skeletons in BB II at

Hasanlu that were wearing lion pins on the day of the destruction? As
reported above and earlier,17 lion pins were recovered,
associated, in groups of one, two, or three, with the skeletons of individuals
killed within the building In one instance three pins were found on a
skeleton.
Eighteen skeletons were recovered unadorned, another 28 are recorded as
not having lion pins, for a total of 46. And of the remaining 16, we have seen
that for 13 we can form no certain opinion about presence of lion pins and
the number worn. These figures are the evidence, the data, from which to
move and make or refuse to draw conclusions.
That lion pins were indeed worn by a good number of the 62 individuals
who died in BB II is manifested by the large number excavated, and is not
a contested issue. All the clusters or units of pins found on the ground,
near a skeleton or not, must surely have fallen from individuals at the
time of BB IIs collapsewhich would explain why many skeletons were
recovered unadorned. But the substantive issue before us is that we are not
able to reconstruct how the 69 pins originally were apportioned among 62
skeletonssome of which were (see below) enemy soldiers. Put another
way, archaeologists cannot know (and should not be tempted to estimate)
how many of the 62 individuals were adorned with jewellery, and how many
were not, and how many wore lion pins, how many did not. Also, how many
of them (adorned or not) belonged to, were stationed in BB II in some official
manner, or had fled there to escape danger.
Concerning numbers alleged to have been worn by age groups, of the
three skeletons with in situ lion pins, skeleton 263, one young adult, wore
3 lion pins; skeleton 255, another young adult, skeleton 255, had one
and possibly two more; and skeleton 256, a child, wore two lion pins. This
evidence modifies Marcus claims about age/number correlation, and is the
only objective evidence we possess about lion pin loci on skeletons.
Marcus claim that lion pins were pinned on the upper body is correct.
The lion pin of skeleton 255as shown in the photograph (Fig. 3) could have
been displaced from the chest or shoulder. Because of their weight, and
more so the thickness of the long pin, they probably pierced a thick woolen
garment. Recorded in one instance were three lion pins ranged one above
the other (skeleton 263; Fig. 2). It appears that at least two were originally
17 Muscarella 1988, p. 42.

positioned one over the other (probably horizontally) across the body, at
the stomach and chest areas, and the third either at the upper chest, or the
neck, or on one shoulderbut were displaced when the individual fell and
was crushed.
We are also informed that the excavated skeletons could not be sexed
by methods of physical anthropology, their bones being too crushed by
fallen building debris.18
This is the understanding I recall from excavation at the site in 1960 and 1962,
and which is well-attested in field notes and photographs. Nevertheless,
aside from skeleton 135, it is reported that two (bona fide) lion pin-bearing
skeletons, skeleton 255 (see below) and 263,19 were sexed as females by a
physical anthropologist, Dr. Janet Monge, on the basis of field photographs
(i.e., here Fig. 2). Disregarding Monges warning (reported in her note 20)
that photo sexing is highly speculative, Marcus requests that we tenta-
tively accept them as females. The information I have been provided is this:
a physical anthropologist has concluded with 75 % confidence that skele-
ton 263 is a female, based on the skull and part of the pelvis. Unfortunately
this report was followed by another, stating that in fact, and for reasons
unknown, there are two skeletons labeled skeleton 263, and it is impossi-
ble to determine which is the one with the lion pinsor if that skeleton is
in fact the one in the University Museum.20 Hence, the skeleton 263 pho-
tographed in situ with lion pins remains unsexed.
There is also a tendentious shift concerning interpretation and alleged
excavated data from the claim made in the 1993 paper that children as
young as six or seven wore lion pins, to that of the 1995 paper that females,
including girls as young as six or seven wore lion pins.21 Children transmo-
grified into girls, nothing about boys (and skeleton 256, a child, has to my
knowledge not been sexed).
In 1993, figure 14, this notion was carried further. It is a drawing designed
to show the location of personal ornaments and alleges to demonstrate not
only that females wore lion pins but precisely where on the bodyand why.
18 Marcus 1933, pp. 159, 170, and notes 19, 20, 23; also in 1994, p. 4, and 1995, p. 2502.
19 Marcus 1993, p. 170. I have no knowledge how my text in Muscarella 1989, p. 32 stated
that females and children wore lion pins, and citing fig. 14 as an example: but in fact, the
caption here was not written by me but by the editor. In any event, the present paper
confronts these issues.
20 These reports from Dr. Janet Monge reached me over several months from September
to December 2001 via e-mail; I thank her for her help.

21 Marcus 1993, p. 164; 1995, p. 2503.
292 chapter eight
The caption says Reconstruction of Skeleton 135 as female , but whose sex
we are simultaneously informed is not based on any physical evidence,22 for
it was very badly preserved. Listed in this skeletons field report sheet as its
furnishings were bracelets, a bowl and spoon, iron spears, a helmet, of which
more below,23 and three lion pins, whose positions whether on or near the
skeleton were not recorded. In the drawing three lion pins are depicted worn
on both shoulders and, nota bene, none are depicted worn on the upper body.
The shoulder positions are rationalised as a possible [sic] arrangement
(some explanation for this females preference for asymmetry would be
useful, or how sheand all the other BB II females?choose two pins for
the right, one for the left shoulder). To document the drawings verisimili-
tude it is claimed (in note 23) that the execution is based on the available
field photographsnot one of which is published to support this alleged
fact: because none exist. Her own figure 2 shows precisely where lion pins
were worn, which is not three across the two shoulders. Skeleton 135 had no
finger rings or pins recorded in the field report, but there were two snake
headed bracelets, one with raised beads, and also a flat band form not shown
in figure 14 (several snake headed bracelets were also reported associated
with the skeleton of an infant, skeleton 148).
Among other adornments drawn on the akimbo female is a necklace
amulet composed of a fish on a chain.24 This amulet appears neither in the
inventory list in the furniture section of the skeleton sheet, nor on the field
object card as from skeleton 135. It was listed as from that skeleton in a
Register book, which information was passed to the Metropolitan Museum
of Art when the object came to reside there as its share of the excavation
division.25 The fact is, no one knows exactly where the amulet was recovered,
and it (along with the garment pins, and finger rings) should not have been
placed on the fleshed body of figure 14.
In wishing to demonstrate how females wore lion pins, the selection of
skeleton 135 was an unfortunate choiceit cannot be sexed, and the mor-
tuary furniture listing of the helmet and spears could indeed have allowed
it to be interpreted as a male. Indeed, where in figure 14 are (his?) helmet
and spears listed in the mortuary furniture, which listings are otherwise
accepted by Marcus? In fact, however examination of the field record reveals
22 Marcus 1993, p. 174, note 2; infra.

23 Muscarella 1988, p. 50, No. 60; 1989, fig. 9a.
24 Muscarella 1988, No. 41.
25 Which is why in Muscarella 1988, p. 41, I said it was from the same skeleton that had a
helmetibid., p. 50, which also did not derive from skl 135 (see below and note 26).
that there was no helmet on or near skeleton 135. A sketch of their loci shows
that the helmet was recovered at some distance from the skeleton;26 no
other artifact, including the spears, is placed in the sketchand their exact
loci too remain unknown. It follows that one may not use the associated
weapons to justify the claim that skeleton 135 is male, but neither may it be
declared a female: it remains anthropologically unsexed. Figure 14 is a ten-
dentious leap, deconstructing the ancient excavated and recorded evidence;
it is an artifact manufactured in the modern world.27
If sexing cannot be accomplished anthropologically, the BB II skeletons
can be sexed however speculative [sic] the exercise.28 by examining juxta-
posed, associated, or actually worn artifacts (which are differently defined
situations) from the skeletons and comparing them to those from the ceme-
tery, where it was generally possible to determine sex (but where no lion pins
were recovered). Fine; but the presented conclusions raise doubts in some
cases. For example, short garment pins worn on the shoulders to Marcus
signify females because she claims more females than males wore them in
the cemetery, and also throughout western Iran. But this equation obscures
the reality that males also wore them in the cemetery (at least one example),
and in BB II. An adult male, 3040 years old (skeleton 260; anthropologically
sexed) from BB II had two straight pins recovered at neck, one of which
was 8cm in lengthshortand a third close by. Further, a male is rep-
resented with a pin placed seemingly on one shoulder on the Hasanlu gold
beaker.29 Marcus genders four skeletons [135 (see above), 139, 140, 255] as
26 Correct therefore Muscarella 1988, p. 50.

27 Marcus invokes the mantra feminist art history (1993, p. 171) to empower one to
reveal and demonstrate the hitherto hidden/suppressed non-marginal roles of women in
antiquity. What has been demonstrated here is that one need merely invoke this gendered
label to become self-licensed to ignore evidence acquired in the field, and thereby to become
genderly creative.
28 Marcus 1993, p. 170. It is unfortunate that one cannot check the essentially secondary
citations in her (in-part) overlapping articles because the majority of the Hasanlu artifacts
(those in the many burials for example) remains unpublished in a systematic manner 40
years after excavation.
29 Marcus 1994, pp. 4, 7, fig. 7, for garment pins from the cemetery. On the Hasanlu gold
beaker two females wear pins with the points protruding up from the garment; and one
seated/squatting male wears a pinonly one is shown, on the proper right shoulder (Porada
1965, pp. 9899, figs. 633, 64); other males and females represented here do not wear pins.
The males pin has a large rounded top with the pin tucked down into the garment; in
size it appears to be large. In 1993, note 23, Marcus cites this gold beaker male as visual
comparanda for her creation of fig. 14but how, whether for straight or lion pins is not
explained (I cannot see any parallels either way, and the garment pins are not in the same
positions). To explain why a male is a model for a female in pin adornment, in 1994, p. 12,
294 chapter eight
female because they have short garment pins at the shoulders.30 For skele-
ton 139 there is no mention of a garment pin in the mortuary furniture listing
(the skeleton, in poor condition, was not completely excavated); for skele-
ton 255 a pin is listed as near arm. She does not mention that male skeleton
260 (above) wore garment pins, at least one of which was short (a straight
pin was recovered juxtaposed to skeleton 263 but where was not recorded).
Further, in other venues31 it is large pins, extremely long, not short ones,
that she associates with females (see below).32
Negative evidence is introduced to support the claim that
none of the skeletons with lion pins can be associated with male-gendered
artifacts
which includes belts, armbands and weapons.33 But this conclusion ignores
her normal methodology that accepts (where convenient) all the listed mor-
tuary furniture, including lion pins, as associated with, that is, belonging
to the assigned skeleton. For example, the spears and the helmet men-
tioned above in the mortuary furniture with skeleton 135 are overlooked,
whereas the lion pins listed in the same furniture inventory are emphasised
as manifestly belonging to a female skeleton. Overlooked is a dagger and a
knife recovered centimeters away from skeleton 263, which Marcus sexes
as female;34 in Fig. 2 the top of the dagger is seen at the lower right. That
this juxtaposition could have been fortuitous, the weapons having fallen
from another doomed individual, is surely possiblebut we cannot know.
Nevertheless, here is an example of male-gendered artifacts that are indeed
associated with a skeleton wearing lion pins. Skeleton 152, an adult, had
she casually states that the male wearing the pin on the gold beaker suggests [to Marcus]
that certain cult officials were gendered female merely because he wears a shoulder pin!
(Whether to her the pin-wearing males recovered at Hasanlu are also to be understood as
cross-dressers is not revealed to us.) This is bald politically correct behavior.
30 Marcus 1993, p. 170; for skl 135s pin see Marcus 1994, p. 7, fig. 7, E.
31 Marcus 1994, p. 11; 1966, pp. 4647, but compare this to p. 49 where both males and
females both wore short pins; see also note 32.

32 Note that the Surkh Dum pin worn by a female in her 1993 fig. 12 (also 1994, fig. 11; 1995,
p. 50, fig. 24) is not short, as she claims, but fairly large; as is that worn by the male on the
gold beaker (see note 29). On the former, a female, the pin points up, on the latter, down; see
also an unexcavated but western Iranian, probably Luristan, disc pin where what is probably
a female wears two very large pins with the pins up (Porada 1965, p. 88, fig. 60). Thus, long
pins were worn by males and females in western Iran, and the pin (at least for females) could
be positioned up or down indicating that neither the size nor position of garment pins was
gendered; see also note 31.
33 Marcus 1993, pp. 170, 174, and note 19.
34 Visible in Muscarella 1989, p. 32, fig. 14.
three lion pins listed as part of its furniture, and an iron sword over its leg;
skeleton 157, also an adult, had a lion pin listed in its furniture inventory, and
an iron spear point near a leg. Marcus should have brought the associated
weapons into the gender discussion of these three skeletons, if only because
her methodology of sexing by means of associated material logically should
have indicated to her that they could be sexed as males.
Sexing artifacts is a growing research strategy of archaeologists; all are
concerned with investigating gender preferences and interpreting indige-
nous culture-controlled rules, which can be different from culture to cul-
ture.35 Warranting investigation are the cultural rules that determined
gender-specific forms of jewellery: anklets, bracelets, earrings, necklaces
and lion pins. Evidence for gender-shared and gender-neutral categories
of artifacts at Hasanlu and elsewhere is presented by Marcus;36 anklets are
included in the latter category and she notes
there is absolutely no reason to assume that the lion pins were gender-
specific,
a viable conclusion. But, then, did men and women equally wear lion pins,
and what of children?
Unsexed skeleton 255, a young adult, wore one dangle earring, anklets,
bracelets, a necklace and lion pins. Skeleton 256, a child, wore dangle ear-
rings, a bracelet, and at least one lion pin. Skeleton 263 a young adult
(female?), wore a necklace of amber beads, one garment pin, and three lion
pins. Now, Marcus claims that dangle earrings like those worn by skele-
tons 255 and 256 were recovered exclusively with females in the ceme-
tery burials. Since this evidence is not published, it cannot be verified,37
but if this earring-specific gendering is eventually demonstrateddangle
earrings equal female (and also a number of earrings, still to be identified,
equal males)then a child and a young person who wore lion pins may
have been females.38 Unless one accepts the weapons close to, associated
35 Thus, in the Near East from the third millennium down through the Achaemenian
period, only royal males were represented with parasols, whereas in Classical Greece only
women and openly homosexual males were represented holding them. See my Parasols in
the Ancient Near East. 1999 SOURCE XVIII, 2: 1 ff.
36 Marcus 1993, p. 170; 1995, p. 2500, and fig. 14.
37 What is published are drawings of earrings that we are told come from female burials
at Hasanlu: Marcus 1995, p. 2501, fig. 16.

38 A good example of the problem of gendering jewellery, in this case earrings, is demon-
strated nearly by the Ashurbanipal garden scene relief: the king and queen wear the very
same earring form; the females to the left of the queen and the males (eunuchs?) to the kings
296 chapter eight
with, skeleton 263 as evidence it is a male, one cannot otherwise yet iden-
tify a skeleton that manifestly wore lion pins as male (as noted the sex of
skeleton 263 remains ambiguous). It is probably correct to state that archae-
ologists have not demonstrated that lion pins were or were not gendered at
Hasanluonly that it seems some females wore them. What appears cer-
tain is that whether gendered or not, they functioned as a marker exclusive
to Hasanlu.39 Oddly, especially given her figure 14, Marcus presents an anti-
gendering conclusion:
In the end however, in the absence of good physical criteria, the biological sex
of the skeletons with lion pins remains uncertain.40
In this context it would be valuable if one could determine not only if
straight pins were gendered by size, that is, whether or not males and
females wore different sized straight pins (above; more investigation is
needed), but relevant here, whether this situation equally existed for lion
pins. No one has ever introduced the subject of size differences among the
lion protomes and what this might suggest.41 An incomplete investigation
on the lengths of the lions yields a considerable variety: 3.6; 3.8; 4.2, 3; 4.8;
5.0; 5.2, 5.4, 5.6, 5.7, 5.8; 6.2; 6.4; 6.5; 7.1 and 9.3 cm. The longest is nearly
three times the length of the smallest. The two lions deriving from skele-
ton 256 (child) are 4.8 and 5.0cm in length; the three with skeleton 263
(young adult) are 5.4, 5.8, and 6.2cm in length. The three lions associated
with skeleton 255 (young adult) are 3.6, 4.3, and (the one from the humerus)
5.2cm in length. The measurements are not uniform, and the young adults
pins do not closely conform in size, one being among the smallest in size
(3.6cm). Two proposals deserve consideration: that size differentials reflect
hierarchical differences among a select (still to be defined) selfreferential
group (religious devotees?), whether gendered or not; or, size relates to
right wear forms different from the royal couple as well as from each other. In the attendants
case, gendering may inform the difference, in the royal situation high rank may inform the
wearing of the same earring form. If so, we may have before us two separate (and simulta-
neously depicted) functioning levels of artifact selection: gender and rank. See Barnett 1975,
pl. 170.
39 Marcus 1933, p. 165, plays with this idea; see below in my text. I find it of interest to
note the formal parallel of the lion pins to a metal foundation peg with a recumbent lion
terminal excavated at Bismya (Banks 1912, p. 237; I know this peg from a discussion with Jean
Evans): is there an underlying ideological issue here? And how does the Tishatal/Urkish lion
foundation peg, here with an upright lion (Muscarella 1988, No. 495), fit into this ideology?
40 Marcus 1993, p. 170, note 19.
41 Marcus 1995, pp. 2497, reference is made that they are huge, heavy, see also p. 2503. In
Muscarella 1988, I neglected to give measurements for the lions per se.
gender and age, large pins for males, medium ones for youths and women,
and small ones for children. This latter proposal is not mutually exclusive
from the first. But as we have seen, there is insufficient data to speak either
way about size differentials and age, but the limited evidence suggests that
size is not related to age. However, because of skeleton 256, we do know that
some children wore lion pins at Hasanlu.
The perception that women wore lion pins triggered Marcus to construct
a significant cultural interpretation, one founded under a canopy of theory:42
It is tempting [!] to think that competition over resources
and Urartian and Assyrian expansion led to a situation where lion pins
created a need to stress unambiguous identities,
and that they
may be seen as part of a broader cultural desire to create the illusion of an
armed and well-defended society an actual military presence as well as
artistic representations of military might
This is especially the case, inasmuch as the lion pins
resemble swords and daggers with lion hilts (cf. fig. 10)
In fact, Marcuss figure 10 shows only a non-human headed hilt that is not
a lion head (it looks like a snake), and, in any event, the weapon depicted
here derives from Sialk, not Hasanlu. Further, the lion pins
carried a more overt [than straight pins!] message about military strength

and
especially compelling about the Hasanlu situation is the notion [to the
author] that body ornaments and hence human bodies played an active role

in not only announcing social rank, but also
promoting an ideology of military strength and power 43
And last but not least, women wore lion pins because their
bodies complemented or reinforced traditional male signals of military
strength [and] adorned females would have been an essential part of the
42 Marcus 1993, p. 168; an attenuated version of the outside threat and lion pin reaction to
it appears in 1995, pp. 24972498.

43 Marcus 1993, pp. 168170; see also Marcus 1996, p. 49.
298 chapter eight
iconography of power and actively involved along with other members of

the society [i.e. males] in promoting a cultural ideology of military might 44
The above ideology was presented in the 1993 publication. However, in the
1995 paper, a different approach was presented (but with no explanation for
the change).45 Here, there is nothing about womens military ideology and
support of their men folk. Rather, we get a simpler interpretation that the
pins demonstrate that females had
formal, ceremonial roles in the temple or active roles elsewhere at the
site
or were members of a palace elite with a ritual or royal dress . But
her 1996 paper,46 harks back to 1993 again, with all the military ideology
presented for both straight and lion pins, here again with no explanation
about changing, shifting, floating conclusions presented over four years.
The question whether the pins represented a particular group identity
marker, of elites, or palace or temple personnel, or a religious or secular
rank or role marker, is to me the position to pursue. It makes sense and
is viable, whether the pins were gendered or not, and whether BB II was
a palace or a temple.47 Marcus 1995 suggestion48 that the lion may be an
indication of some service to a divinity whose attribute is a lion makes
sense. But I suggest that her general conclusions were written to share the
range of revelations made possible by invoking theory, which power discov-
ers hitherto hidden or suppressed (alas, by both men and women scholars)
knowledge of ancient womens real activities in society. This theory toler-
ates a situation that allowscompels?us to accept notions that women
at Hasanlu were more empowered, integrated, and independent than one
hitherto thought. It reveals the real archaeology and history of Hasanlu that
had not hitherto been perceived: militant women, arms akimbo, stand with
equality alongside their men folk.49 No marginalised women they. But absent
44 Marcus 1993, p. 172; note that in 1994, p. 10, she presented the same interpretation for
women wearing straight garment pins.

45 Marcus 1995, p. 2503this latter position is touched upon in 1993, p. 172, but aban-
doned.
46 Marcus 1996, pp. 42, 46, 48, 49.
47 In Marcus 1993, p. 165 BB II is a palace. In 1995, pp. 2496, 2502, BB IIs identity is shifted
and it is called a temple. The function(s) of BB II is complex and is still not satisfactorily
explained.
48 Marcus 1995, p. 2503.
49 Marcus fig. 14 seems to illustrate the sense of the Western male labor song, here of
course transgendered: When you see me coming, better step aside/a lot of men didnt, a lot
of men died. Whatever the case, women died alongside men at Hasanlu.
theory, no evidence from the archaeological or textual record from Hasanlu

or any site in the Near East supports this scenario. It is not in harmony, it does
not resonate with the ancient evidence, and it is not compelling or tempt-
ing.50
A note about the 62 dead recovered in BB II is relevant. Most of the
skeletons were recovered in the northern section of the great pillared hall of
BB II, in the entry Portico (Room 2), in the entry doorway to the pillared hall,
and in areas adjacent to this doorway area on both east and west.51 Among
the dead were adults, including young adults, children (11) and infants (3).
One ponders whether they were destroyed while attempting to flee from
the burning building through the entrance/exit (as suggested by several
members of the Hasanlu team), or the very opposite,52 fleeing there for
safety, but alas, failing in their goal. What is certain is that whatever the
historical reality for the concentration of 62 dissimilar individuals in BB II,
enemy soldiers were among them: they had penetrated into the pillared
hall. Two crested helmets, spears, maces, swords and daggers recovered on
the floor (distinguished from the mass of spears recovered in the fill from
the second storey collapse), or under or close to some skeletons, could have
belonged to the attacked and/or the attackers, and archaeologists would not
find it easy to distinguish among them. But weapons that most certainly did
not belong to the defenders are recorded on the skeleton record sheets made
in the field: blade through head of skeleton 125, a child; 2 spears through
waist of skeleton 128, an adult; and iron weapon through L hand of skeleton
156, a child. Enemy soldiers (probably Urartians) and local inhabitants died
together in BB II when the building collapsed, but the soldiers were not
wearing lion pins.
Bibliography
Banks, E.J.
1912 Bismya: or The Lost City of Adab. New York and London: G.P. Putnams Sons.
Barnett, R.D.
1975 Assyrian Sculpture in the British Museum. Toronto: McClelland and Stewart.
50 I cannot refrain from quoting Marcus (1995, p. 2488): The Hasanlu material, however,
lends itself to an evaluation of modern preconceptions and ideological agendas applied to

the examination of artistic remains that do not include written texts.
51 Muscarella 1989, p. 35, fig. 20.
52 In Marcus 1993, p. 163, the victims were fleeing BB II; in 1995, p. 2499 they are also fleeing,
but on p. 2502 they escape for safety into BB II.

300 chapter eight
Dyson, R.H.
1964 In the city of the golden bowl. Illustrated London News 12 (September):
372374.
Marcus, M.
1993 Incorporating the body: Adornment, gender, and social identity in ancient
Iran. Cambridge Archaeological Journal 3 (2): 157178.
1994 Dressed to kill: women and pins in early Iran. Oxford Art Journal 17(2): 3
15.
1995 Art and ideology in ancient western Asia, in Civilizations of the Ancient
Near East, edited by J.M. Sasson, J. Baines, G. Beckman and K.S. Rubinson,
pp. 24872505. New York: Scribners.
1996 Sex and politics of female adornment in pre-Achaemenid Iran (1000800
B.C), in Sexuality in Ancient Art: Near East, Egypt, Greece and Italy, edited
by N. Boymel Kampen, pp. 4154. Cambridge: Cambridge University Press.
Muscarella, O.W.
1988 Bronze and Iron. New York: Metropolitan Museum of Art.
1989 Warfare at Hasanlu in the late 9th Century B.C.. Expedition 32 (2/3): 2436.
Porada, E.
1965 The Art of Ancient Iran: The Art of Pre-Islamic Times, London: Methuen.
Figure 1. Hasanlu lion pins: Upper pin: 60983, MMA 61.100.10.

9.3cm; Lower pin: 60560, MMA 61100.15, L. ca. 3.9cm.
302 chapter eight
Figure 2. Hasanlu Tepe: Skeleton 263, BB II, top

of dagger is seen below the skeletons right knee.
Figure 3. Hasanlu Tepe: Skeleton 255, BB II; photograph courtesy of

University of Pennsylvania Museum of Archaeology and Anthropology.
304 chapter eight
Figure 4. Hasanlu Tepe: Skeleton 157, BB II; photograph courtesy of

University of Pennsylvania Museum of Archaeology and Anthropology.
chapter nine
THE EXCAVATION OF HASANLU:

AN ARCHAEOLOGICAL EVALUATION*
Abstract: The site of Hasanlu in northwestern Iran was excavated in 10 campaigns

between 1956 and 1974. Publications of the excavations by the director Robert
H. Dyson, Jr. and by expedition staff members have appeared in many academic
journals and in a number of other venues up to the present, including several mono-
graphs. They amount to more than 90 works, but only one final, synthetic report has
ever appeared. Much of Hasanlu still remains underpublished and unpublished.
Further, a review of the publications that have appeared across the past 50 years
reveals an inadequate and inconsistent, sometimes confusing, documentation of
the various cultural levels uncovered with the artifacts, architecture, burials. and
chronology. This study articulates these issues by confronting and evaluating these
publications in the sequence of their appearance and discusses how this record of
the site affects our knowledge of one of the most important sites in the Near East,
as well as the archaeology of northwestern Iran.
Introduction
Archaeologists in search of in-depth information, artifact data, or a sys-

tematically presented diachronic survey of the chronologies and cultural
histories of the various levels uncovered at the site of Hasanlu in northwest-
ern Iran from 1957 to 1974 soon become aware that this information does
not exist in a convenient and coherent form. Information published about
the diverse cultural levels, stratigraphical changes, architectural details, arti-
fact records, and burial data is broadly scattered, rare, or nonexistent; and
obtaining an articulated archaeological broad sense of several important
cultural periods uncovered is difficult or impossible. Much of the infor-
mation published is presented in general terms and scattered over many
* This article originally appeared as The Excavation of Hasanlu: An Archaeological
Evaluation, Bulletin of the American Schools of Oriental Research 342 (2006): 6994.
This article is the revised version of a lecture delivered at the Albright Institute of
Archaeological Research in Jerusalem in March 2006, when the author was the sixth Trude
Dothan Lecturer in Ancient Near Eastern Studies. This series, which also includes lectures at
Al-Quds and Hebrew Universities, was sponsored by the Albright Institute and endowed by
the Dorot Foundation.
306 chapter nine
venues; contradictory data, conclusions, and claims are often not addressed
or explained in subsequent publications. Hasanlu is a site about which it
can be said that too much has been published, and at the same time that
not enough has been published. To some extent, this state of affairs has
been on my mind for years, especially because of my long involvement with
the Hasanlu excavations and Iranian Iron Age archaeology. But only when I
first systematically reviewed and indexed the contents of the complete pub-
lication corpus in its published sequence did the full extent and range of
the problem fully surface. In addition to issues I already knew or sensed,
other significant onesconcerning the prevailing, characteristic manner
and method of reporting information and datawere exposed in the course
of this research.
The present studyan analysis and critique of the publications and how
the site was publishedwas generated precisely because of this review.
The questions I asked were these: What data and information have been
published and what have not? In what manner have they been published?
What do archaeologists know and not know about Hasanlu? And why?
Publication History
The first difficulty encountered is documenting the sites publication his-

tory, and then its information-sharing strategy. In this study, I cite primarily
the publications written both by the Hasanlu director, R.H. Dyson, Jr., and
the excavation staff, including those invited to contribute special studies,
but not the many secondary studies on Hasanlu written by other schol-
ars, viz. Edith Porada, Roman Ghirshman, Inna Medvedskaya, and so forth.
Obviously the Hasanlu researcher must also consult them, for they amount
to more than 40 publications. But they are not relevant to the thrust of
this study, which focuses on the primary Hasanlu publications. For more
than four decades-from 1957 to the present-some 89 articles, about 47 by
the director,1 have appeared in at least 20 different journals and in numer-
ous other venues. These comprise brief field reports, summaries of field-
work, discussions of specific artifact types and architecture, or general
cultural matters. The journals include The Illustrated London News, Expe-
dition, Archiv fr Orientforschung, Art Bulletin, Iran, American Journal of
1 Two articles are essentially the same, Dyson 1962 and Dyson 1966; several are joint
articles; and I suggest that Dyson 1999a and 1999b should have been published as one article.
the excavation of hasanlu: an archaeological evaluation 307
Archaeology, Archaeology, Bulletin of the Canadian Society for Mesopotamian

Studies, Explorers Journal, Iranica Antiqua, Journal of Near Eastern Studies,
Cambridge Archaeological Journal, University Museum Bulletin, Oxford Art
Journal, Bulletin of the Metropolitan Museum of Art, Archeologia Viva, Natural
History, Journal of Glass Studies, Science, and Anatolian Studies. Other pub-
lished venues include symposium volumes, publications of collected arti-
cles, festschrifts, encyclopedias, and museum publications, viz. Encyclopae-
dia Iranica, Bronze and Iron (Muscarella 1988), The Oxford Encyclopedia of
Archaeology in the Near East, Civilizations of the Ancient Near East, and A
Survey of Persian Art. In addition to the articles, there are five monographs
dealing with specific levels or categories of artifacts: I. Winter on a breast-
plate and some horse gear (1980); Muscarella on the ivory corpus (1980);
M. Marcus on the seal corpus (1996a); T. Rathbun on skeletal identification
(1972); and M. Danti on the Islamic level, Period I; also one Ph.D. dissertation
(Young 1963).
But there exists not one comprehensive final report in the publication
list except for the Islamic period (Danti 2004)no central venue to which
one may refer for a unified synthetic analysis of the diachronic changes of
the cultures uncovered, for a central unified discussion of artifacts and archi-
tecture, for corrections and analyses of errors and mistakes, and clarification
relating to the many changes of opinions or interpretations over time. This
situation existed during the 10 field campaigns (19561974) begun exactly a
half-century ago and continues to the present, after the termination of exca-
vations three decades ago.
Each individual researcher is obliged to develop a strategy related to
acquiring both general and precise knowledge of the site. A great deal of
time must first be expended in recording the scores of publications and then
searching for them in several libraries-if they are available. Even Hasanlu
staff members will have many but not all (I did not) of the publications.
A bibliography of publications relating directly or tangentially to Hasanlu
Project excavations (see n. 2 below) or to related issues was published in
Levine and Young *1977: 399405; see also Dyson and Muscarella 1989:
22, n. 1, and 2627. Those relating specifically to the Hasanlu excavations
are of course included, but they must painstakingly be sought out, and
obviously more recent publications must be added. I therefore compiled a
(I trust up-to-date) full bibliography that time-lists publications that relate
solely to Hasanlu excavation matters. To facilitate both bibliographical and
internal scholarly research, I sorted the bibliography unconventionally, in
the order that works were published and read by scholars. I also include
here a separate bibliography of related publications cited in this study,
308 chapter nine
organized in the traditional way. Text citations to works in this latter bibli-
ography are indicated with an asterisk (*).
Researchers must then organize a system to record data and information
systematically in the chronologically published sequence on specific or gen-
eral topics included in the many publications. It is frustrating work. Reports
on a specific years campaign are uneven, often brief, at times merely a para-
graph or a page (see the Addendum below). Sometimes it is unclear which
campaign season is being discussed. (For example, Dyson 1960c is in part a
report on the 1959 campaign, but this is not mentioned. I recognized this
only because the Bead, Artisans, and South Houses were excavated that
year.) And it soon becomes apparent that, along with new information scat-
tered across the publications, there is also a repetition of the same topics,
and photographs of the same artifacts are presented-some four to six times
in as many publications-while many others remain unpublished. Confusion
also arises from the character and minimal quantity of the information pre-
sented, from the many changes in factual statements and interpretations
presented across time with no back referencing or explanation, and from
the methods chosen to reveal the excavation results from each campaign.
To communicate and comment on these matters does not result in an
easy read, although this is not my design. Rather, I would argue, the difficulty
comes precisely from the very nature of the issues I address. And it is a major
aim of this paper to aid scholars to organize and facilitate future research on
this remarkable site.
Excavation History
The record of the early investigations at Hasanlu is given in Hakemi and Rad
1950 and Dyson 1967: 2951. In 19341935 Hasanlu was commercially dug by
licensed dealers. Stein (1940: 388) reported pits dug here and there for
Jew dealers from Urumyeh. Some of the material recovered was published
by Ghirshman (*1939: 7879, pl. C). The first manifest excavation there was
accomplished by Aurel Stein in 1936, published in 1940. In 1947 Hasanlu
was again commercially dug, by Mahmud Rad, some of which material
was published by Rad and Hakemi in 1950. The second excavation of the
site, a campaign of two months, was conducted in 1949 by Ali Hakemi
(Hakemi and Rad 1950). Finally, beginning with a 10-day survey in 1956
(Dyson 1957: 3739), from 1957 to 1974 nine excavation campaigns were
directed by R.H. Dyson, Jr. The term Hasanlu Project was coined to identify
the strategy of excavating, sometimes simultaneously with Hasanlu, other
mounds in the Solduz Valley, to ascertain additional information about the

areas cultural and political history (Young 1959a: 4; Dyson 1967: 2951); aside
from Hasanlu, seven other sites were investigated.2 But Hasanlu is the best
known and, in fact, one of the most important sites recovered from ancient
Iran, which per se justifies an evaluation.
Synopsis and Evaluation of the Published Data
[T]housands of objects were recovered (Dyson 1997: 480); more than

7000 objects were recovered (Dyson and Muscarella 1989: 1); Over 7000
artifacts were identified (Dyson 2003: 43) from the levels excavated at
Hasanlu. The last reference (repeating essentially the same listing in 1997)
states that they include a wide range of utensils, weapons, jewelry, deco-
rative wall tiles, metal and ceramic vessels, horse gear, seals and sealings,
and so on [italics added]. Materials represented include iron, bronze, gold,
silver, antimony, shell, ivory, bone, amber, glass, wood, and stone. Omitted
in this inventory is mention that the great majority of objects (aside from
ivories and seals), along with their loci, remain unpublished, while some
types are published in part, ad hoc, and within many venues. The artifacts
derive both from the mound itself, called the citadel, about 200 m in diam-
eter, and the nearby cemetery.
For the architecture on the citadel, the publication record is better for
certain periods than others. But the data are spread over many venues
over decades, and to track down information about a particular structure
or a complete plan of a particular cultural level is frustrating and time
consuming. One also encounters different architectural plans of the site.
For example, compare Muscarella 1980: 3 with Muscarella 1988: 17, fig. 2,
which is the same as Muscarella 1995: 992, fig. 8: the latter plan is accurate.
2 Dinkha Tepe: a Bronze and Iron Age site (the former cultural area directed by Dyson, the
latter by Muscarella (Muscarella * 1968; * 1974)); Agrab Tepe: directed by T. C, Young, Jr. and
Muscarella (Muscarella * 1973); Qalatgah (Muscarella * 1971a; see n. 19 below): the latter two
are both Urartian sites (along with Hasanlu III B); S Girdan tumuli: directed by Muscarella
(* 1969; * 1971b; * 2003), a late fourth millennium b.c. Maikop site, The Neolithic sites of Pisdeli,
Haji Firuz, and Dalma Tepe were excavated in 1958, 1959, 1961, and 1968, directed by Charles
Burney, T.C. Young, Jr. and M. Voigt. Ziwiye was added to the Projects goals; it was surveyed
and excavated by Dyson for several weeks in 1964; it remains incompletely published (Dyson
1965b: 205206. and 1972: 5051). See also Dyson in Dyson and Muscarella 1989: 22, n. 1. In
addition to Hasanlu, the sites of Dinkha Tepe, S Girdan, and Haji Firuz were first recorded
by Stein: he was the guide, and the Hasanlu Project followed in his footsteps.
310 chapter nine
Compare also Dyson 1997: 479, fig. 1, and 1989a: 10, fig. 11 (see fig. 2, this
article) with Dyson 1989b: 115, fig. 10 (see fig. 3, this article); the latter is more
accurate. And various pieces of information about specific structures are
presented in different publications-with rare reference to earlier statements
or interpretations.3
An extramural cemetery was excavated below and adjacent to the citadel
mound, in the area called the Outer Town (Dyson 1958a: 27; Young 1963:
2223; Dyson 1989b: 107); and some burials were also recovered on the
citadel itself (Dyson 1965a: 158). Although the cemetery was a major source
of juxtaposed pottery and artifacts, no definitive information about the
precise number of burials recovered, their period-by-period breakdown, or
group contents have been published. In Muscarella 1966: 134, I claimed
there are over 150 graves from Periods II to V (see also n. 11). As late as
January 2005, in response to email enquiries, M. Danti told me that the
number of burials from all cultural periods recovered is not known (!), that
the figure remains to be tabulated (there are probably scores of burials).
Further, not a single burial from any period has been completely published
with all its burial goods and contexts discussed (see below for specific
details).
The following commentary is a synopsis and evaluation of what I submit
we know, and do not know, from the publication record regarding the
various cultural periods at the site, which are numbered consecutively in
Roman numerals from the top of the mound down, I to VII. Excavation
was controlled by laying out 11 11m squares, leaving a 1 m balk. These were
originally assigned Roman numerals, XL, XLI, etc. (Dyson 1959a: 8; 1966:
fig. 16-3; Young 1959a: 4), and had been laid out as a grid plan; these labels
were later converted to A, B, C, AA, BB, and so forth.
Period VII AC
Labeled the Painted Orange Ware Phase, or Earlier Bronze Age (Danti,
Voigt, and Dyson 2004: 588), this period is referred to inseveral venues. In
Dyson 1958a: 26, fig. 22, a burial is reported and a paint-decorated vessel
therein published-also published in Dyson 1960e: 132133, fig. 7; 1967: 2956,
pl. 1483; 1968: 87. Young (1959b: 65) and Dyson (1967: 2956) discuss pottery,
3 In rare examples, Dyson, in Dyson and Muscarella 1989: 22, n. 1. states that his publica-
tion 1965b is an outdated article, and Dyson 1999a: 136, says fig. 12, is a revised pottery chart
replacing an earlier one, but in neither case does the author give reasons to explain these
comments (see also Hasanlu III B and A issues, below).
and the latters fig. 1027 provides drawings. Dyson and Pigott (1975: 182) noted
that there are three phases in this period and briefly discussed the pottery.
That Early Transcaucasian Culture wares were present here in VII C was
first reported in 2004 (Danti, Voigt, and Dyson 2004: 586,594596, fig. 11:1).
The latter publication is the most extensive report of Period VII published;
included (pp. 588593, 602616) are descriptions and drawings of the pot-
tery wares from the three respective phases; and a section drawing is also
provided (2004: fig. 2). The paint-decorated vessel is mentioned, but not
illustrated (2004: 588), btit no burial is mentioned.
Period VI
Period VI is called the Khabur Ware or button base period, named after
very similar pottery from northern Mesopotamia, or Late Bronze Age
Period VI (Danti, Voigt, and Dyson 2004: 586). This pottery is mentioned
in Dyson 1958a: 27, 30; 1963a: 132; and 1965b: 193195, where three ves-
sels are shown with comparanda (pl. 31). In Dyson * 1973: 703706, the
Bronze Age period is discussed, but Hasanlu VI is barely mentioned; and
Dyson and Pigott 1975: 182 has one paragraph in which a fortification wall
is reported. In Dyson 1989b: 108109, this period is again briefly mentioned,
and here for the first time we learn of a settlement, and that it is consider-
ably larger than Iron Age Hasanlu. This important information was left in
abeyance. To date, no plan of this settlement has been published. A burial
with one occupant is mentioned with its finds, but there are no photos
(Dyson 1958a: 3031). Inadvertently, another Period VI burial containing a
male, female, and child, along with istikhan-shaped vessels, was reported
by Dyson (1964e: 36) but was considered by him to be from Period V (see
below).
Regarding chronology, Dyson 1963b: 33 gives a date to around 1500bc; in
1965b: 193, it is ca. 16001250 bc for the occupation range. Radiocarbon dates
are mentioned in Dyson 1968: 85 and 1977a: 165, as evidence that Period VI
terminated in, respectively, 1400 or 1650bc. Later, in Dyson 1989a: 6, this
same event is presented as having occurred ca. 1450 bc As correctly stated by
Dittmann (*1990: 112), the termination date for Hasanlu VI remains unclear.
We have very little information about Hasanlu VI, as Dittmann (* 1990: 108)
also indicated, noting that Dinkha Tepe, about 15 miles west of Hasanlu,
is presently the prime source for information about the Khabur culture in
northwest Iran.
312 chapter nine
Figure 1. Hasanlu Period V buildings (courtesy of the Hasanlu Project).
Period V
In the Hasanlu literature, Period V has been labeled Iron Age I by Dyson
(1965b: 211, table 2;4 on this terminology, see the chronology discussion
below). Hasanlu V represents quite distinctly the appearance of a new
culture and population in the area (Dyson 1977a: 156, 166; Muscarella 1994:
140), with a new, characteristic monochrome, grey ware totally replacing the
previous painted wares; an extramural cemetery; and structures of different
form overlying the Period VI level, with no visibly eroded surface or erosion
deposit separating them (Dyson 1965b: 195). These buildings lie for the
most part directly below the following Period IV level (Dyson 1977a: 156).
However, no section drawings have ever been published to illustrate the
nature of the VI-V-IV transitions as described in print; but it is manifest.
Young (1963: 40) mentions an upper and a lower phase.
4 Young (1963: 133 and following) introduced the ceramic terminology Early and Late
Western Grey Ware Group, and Late Buff Ware Group, terms continued by him in 1965, to
define the three cultural levels. This tripartite terminology was relabeled Iron I, II, and III
by Dyson (in 1965a), who also employed there a variant of Youngs 1963 terminology (Early
Monochrome Grey Wares), but did not cite its original source; Piller, in Stollner, Siotta,
and Vatandoust * 2004: 315316, recognized this. The terms Iron I, II, and III were used by
Dyson periodically (1973c; 1977a; 1999a: 115; Dyson * 1973: 705707, 711), which we all followed.
For criticisms of the term Iron I employed for Hasanlu, see Muscarella *1974: 79. The most
vigorous criticism of the chronological problems associated with the Iron I and II terms is
Young 1985: 362, n. 1.
Six isolated structures have been excavated and mentioned in several

articles, including Dyson 1963a: 132, and Dyson 1965b: 197. In Dyson 1973c:
3, one building is described; in Dyson 1973a: 303304, and 1989b: 1018, fig. 2,
two restored buildings were published, one the same as Young 2002: 387,
fig. 1 (see fig. 1, this article). A full discussion and plans of all the buildings
were presented in one venue (Dyson 1977a), with their loci indicated on the
mound plan (p. 157: fig. 1; the scale for the structures in figs. 2 and 3 is 5, not
10m). The settlement was quite extensive, widely spaced over the central
area of the mound, but no site plan has been presented (it was difficult
to recover Period V architecture because of the presence of the overlying
Period IV structures; see Young 1963: 39); no artifacts other than ceramics
are mentioned here. In Dyson 1989b: 108, fig. 2, two restored building plans
are published. A few artifacts are mentioned or published: in Dyson 1964e
(3435, fig. 2: 1, pl. 9. 1), a dagger from an unpublished burial is illustrated;
Marcus describes two seals (Marcus 1994a: 910, fig. 1; 1996a: 143145), one
from an unpublished burial, one from the Outer Town. Also, a terracotta
human figurine and a pottery bird effigy vessel (Dyson 1973c: 34, fig. 5) are
reported to derive from the fill over a Period V structure; but in the same
year, in Dyson 1973a: 303 (figure caption), the latter is claimed to be from a
Period V building.
The characteristic pottery forms are the un-bridged spouted vessel (it
remains unclear how many vessels of this form were recovered), a button-
base handled beaker, a bowl with a small raised ridge, and the worm bowl
(see Muscarella * 1968: 192193, fig. 17); many (most?) derive from burials.
Available are scattered brief comments and isolated drawings Of photos;
select vessels are illustrated (sometimes with the same photograph: Dyson
1960c: 129; and 1967: fig. 1483b); see also Dyson 1962: 641, fig. 4; 1963a: 132;
1965b: 195196 (illustrated here are the three characteristic forms from Geoy
Tepe, but none from Hasanlu-see fig. 13); Dyson 1966: fig. 16-4; 1967: 2962,
fig. 1033; Muscarella 1994: 141142). Dyson, in 1964e: 3637, 39, fig. 3, and
*1973: 706, mentioned painted, standard Period VI wares as deriving also
from Period V. Further, Dyson described (1964e: 36; see above) an untyp-
ical Period V burial that contained a male, female, and child, along with
four istikhan-shaped vessels (excavated, I believe, in 1959). Young (1963: 40,
43, pi. 13; and 1965: 57, 70, 72, fig. 8:3, 6, 1114) repeated this information.
Dyson consequently interpreted both wares as documenting some cultural
continuity from VI; others followed (viz., Porada * 1965: 108, 252; Muscarella
* 1974: 48; Medvedskaya *1982: 13, 17, 27, 36, fig. 4; Dandamaev and Lukonin
* 1989: 21; Dittmann *1990: 111, and n. 30). That this interpretation was an
error-indeed, that both ware forms in fact derived from Period VI, as did
314 chapter nine
the burial-was first reported in Muscarella *1968: 195; 1988: 378, n. 1; later in
1994: 141, fig. 12.2; also Young 1985: 373, n. 11.
Only one painted vessel derives from Hasanlu, from a Period V burial
excavated by Stein (1940: 401, fig. 110, pls. 24.3, 31.8). Remarkably, a very sim-
ilar vessel was excavated at Dinkha Tepe (Muscarella * 1974: 3940, figs. 3,
5) from what I designated to be a Dinkha III (contemporary to Hasanlu V)
burial, recovered alongside other contemporary burials in the extramural
cemetery. Dittmann (*1990: 111, n. 33) discusses the Dinkha burial and con-
siders its contents to be from Dinkha III. Whether these vessels represent
strays or heirlooms from Period VI, or were imports, remains unknown.
Pottery remains essentially unpublished (noted by Medvedskaya * 1982:
33). Young (1963: 3943; 1965: 7072) presents a good review. But as Levine
(*1987: 233) stated, one looks to Dinkha Tepe for ceramic knowledge of this
period in northwest Iran, and Young (1985: 366) notes that Dinkha Tepe
remains the type site for this period (see also Piller in Stllner, Slotta, and
Vatandoust * 2004: 315, 317, 325, n. 16; and Muscarella * 1968; 1974). Unpub-
lished also is a corpus of pottery recovered below Burned Building V of
Period IV, which is described as typologically closer to that of period V,
apparently representing transitional wares from Period V to IV (Dyson 1973a:
303). A decade later, Dyson (1983/1984: 303) also records that ceramics exca-
vated in 1972 and 1974 show a gradual evolution of period V forms into those
of period IV with increasing elaboration. Excavated over 30 years ago, these
important late Period Vearly Period IV wares remain unpublished. Such is
the present condition at Hasanlu, where the Iron I grey ware introduction
and sequence were first charted.
The bones of 16 Period V skeletons from cemetery burials are recorded by
T. Rathbun (1972: 5257), but the burials themselves remain unpublished
(as frustratingly noted by Medvedskaya *1982: 33).5 In some publications
(Dyson 1958a: 31; 1960c; 1960e; 1962: 639; 1964e: 36; 1965b: 196; 1966: 417; 1967:
2957), burials and their pottery, and beads, bracelets, and one iron ring are
sometimes cited, but no photos are published. The ring is the only iron
artifact recorded as deriving from Period V; no iron artifact was recovered
from the same cultural period at Dinkha Tepe (Dinkha III; Muscarella * 1974:
38). Inasmuch as Dyson reported that the major remains of Period V come
from the cemetery (1965b: 196), the non-publication of the burials is all
5 In 1966, I made enquiries regarding the number of Period V burials excavated. I was
informed by Maude de Schauensee that there were 10 from this period and that 5 others were
unclear, either from V or IV.
the more unfortunate (Medvedskaya [*1982: 33, 103] understood this). Stein
(1940: 401402, pls. 24, 31, fig. 110) published a grave photo from this period,
and Hakemi and Rad apparently published a few (1950: figs. 8b?, 10?, 14), in
each case together with their pottery. That some graves contained important
artifacts is indicated by a belt buckle reported by Winter (1980: 27) from a
burial.
An unresolved problem concerns the chronology of Period V as presented
in many publications across 30 years. In 1960, Dyson (1960e: 132) gives the
beginning date for the settlement as ca. 1500bc. Full-range estimates are
given in later publications: 1963b: 33, as the end of the thirteenth or begin-
ning of the twelfth century bc; 1965b: 195, ca. 12501000bc; 1967: 2957, as
12001000bc, 1968: 85, 13001000 bc; Dyson *1973: 705, 712713, has it begin-
ning around 1350+/-50 bc; and in (the later written) Dyson 1973c (fig. 5
caption), it is 12001000; in de Schauensee 1988: 45, it is 14501200 bc. In the
latest recordings, Dyson 1989a: 6, it is ca. 14501250 bc, but in 1989b: 107, the
beginning date is shortly after 1500bc, repeated by Young 2002: 386. These
chronological dates for the incipience of V significantly differ and are in con-
flict with one another in their published sequence: 1500, 1250, 1300, 1200,
1200, 135050, 1200, 1450, after 1500a back-and-forth difference of some 300
years (see also below)! And the terminating dates are given several times
as 1100, 1200, and then as 1250, a difference of 150 years. Equally troubling
is that not once are we offered a reflection or explanation regarding the
different chronologies presented over many decades. The same inconsis-
tencies obtain for the 14C dates also given in multiple venues, differing from
one another by hundreds of years, sometimes coinciding with the archaeo-
logical dates given above. These are reported in Dyson 1962: 641; 1964e: 39;
1965b: 197, n. 7; 1966: 420; 1967: 2967; 1968: 85; 1972: 57. In Dyson 1977a: 160,
161, 164, MASCA-corrected dates for several of the V buildings are given as
1300bc, 13001360, and 12201260bc, which suggests an emphasis on the
13001350bc range, but in the same year that dates for Period V were given
as 14501250 bc (1989a: 6, above), Dyson (in Dyson and Muscarella 1989: 8)
says the 14C range is 13501150 bc. Muscarella (1995: 989) places the termina-
tion in the twelfth century.
What are the beginning and end dates for Period V (and concomitantly
the incipient date of Period IV)? Although the published record is con-
fusing, we do know that the culture arrived at Hasanlu sometime after
the Khabur Ware period terminated. If this event occurred shortly there-
after, then Period V would more accurately be designated in archaeolog-
ical terminology as a Late Bronze Age phase (chronologically following
Hasanlu VI). And if such were the case, Hasanlu IV (below) would more
316 chapter nine
accurately be labeled Iron I (see n. 4; and more recently, Piller in Stllner,

Slotta, and Vatandoust *2004: 316).
The cause of the settlements termination has not been discussed, al-
though it is reported once that three successive floors in one building were
badly burned, and ashes were recorded in another building (Dyson 1977a:
160161). Of course, domestic fire or enemy destruction cannot always be
determined (except with Period IV, below).
Period IV
This level was labeled Iron II by Dyson (see above and n. 4); it is the most
extensively published period at Hasanlu (figs. 2, 3). A major and complex
elite settlement was built directly over the ruins of the Period V buildings
(Dyson 1977a: 156); this early level is labeled Period IV C. It is often described
by Dyson as a direct continuity from V (1977a: 166), revealing no apparent
stratigraphic hiatus and the transition [sic] to it from Period V (IV C) may,
in fact, have been less abrupt than presently appears (1965b: 197198); also,
there was a major continuity between periods V and IV [written as VI, a
typographical error], and the IV buildings were superimposed on earlier
construction (1983/1984: 303). The settlement experienced a destructive
fire that terminated Period IV C, dated variously to the mid-11th centurybc
(Dyson 1977b: 550), 1100 bc (Dyson 1989a: 6; Dyson and Voigt 2003: 222), or
the 12th century bc (Young 2002: 386). In some IV C buildings, BB II, IV East,
and V, it was recognized that a rebuilding occurred directly over the original
walls (Dyson 1973c: 2; Dyson and Muscarella 1989: 1); this second building
phase is labeled Period IV B (Dyson 1989b: 111, 114; Dyson and Voigt 2003: 222).
A final destruction ensuing from a violent attack and intense fire is secured
to the end of the ninth century bc, ca. 800 bc, determined by archaeology
and 14C data. That this event was caused by an Urartian penetration into the
Ushnu area seems certain (Dyson 1969: 44; Dyson and Muscarella 1989: 22;
Dyson 1989b: 109; see also Ivantchik *2001: 212, 240, 258, 261, for the ninth-
century chronology). A so-called squatters settlement, IV A, was uncovered
over the IV B ruins (Young 1963: 27; Dyson 1977b: 550; Dyson and Muscarella
1989: 2).
In the first decade of excavation (Dyson 1959a: 14; 1960e: 133; 1963b: 33;
1965b: 197; 1967: 2964; 1968: 85; 1969, 44), the beginning date for the period
was given as ca. 1000bc. That date shifted to 1100 bc in Dyson 1972: 42;
1973a: 303; and 1973c, 1; then earlier, to the mid-12th century bc in Dyson
1977b: 550; in 1989a: 7, it is a hundred years earlier, 1250bc, but in the
same publication (Dyson 1989b: 108), it is the 12th century bc; a few years
Figure 2. Plan of Hasanlu Period IV (courtesy of the Hasanlu Project).

318 chapter nine
Figure 3. Plan of Hasanlu IV (courtesy of the Hasanlu Project).

later (Dyson and Voigt 2003: 221), it is around 1250 bce. Explanations for
these important revisions, a range of 250 years for the incipient date of
Period IV-and concomitantly for the termination of V (see above)were
never presented or confronted. Radiocarbon dates are given in Dyson 1962:
642; 1965b: 197, n. 8; 1966: 420; 1972: 57; each date is different. When did
Period IV commence?
An intense fire that caused the collapse of the buildings and the violent
death of many individuals marks the final destruction (Muscarella 1989: 32,
35, figs. 14, 20). A total (?) of 246 skeletons of men, women, and children
were recovered in all areas of the IV B settlement; a good number had head
wounds from enemy violence, indicating that the fire and subsequent build-
ings collapse occurred in the course of the violent struggle within the site
(Muscarella 1989: 3234, 36, fig. 21). Many skeletons were recovered in the
collapsed buildings, 62 in BB II alone (Dyson [1965b: 202] says over 40 indi-
viduals; Muscarella 1989: 3235, figs. 14, 20; 2004).6 In all publications since
1965, the destroyers are assumed to have been the Urartians, which indeed
best fits the historical scenario (Muscarella *1971a; Dyson and Muscarella
1989: 19, 2022; Dandamaev and Lukonin *1989: 19; Piller in Stllner, Slotta,
and Vatandoust * 2004: 319).
Grey monochrome wares continue, but different forms are introduced.
Pottery is often mentioned, sometimes illustrated in drawings and a few
photographs, and contexts are rarely given (Dyson 1958a: figs. 23, 26, from
burials; 1960e: fig. 4, and from the settlement, figs. 6, 8; Crawford 1961: 87
89; Dyson 1964e: 38, fig. 4; 1967: pl. 1485; 1989b: 113, fig. 7). For pottery from
the earlier excavations, see Hakemi and Rad 1950: figs. 1528, and Stein 1940:
pls. 24, 30, 31 (all mixed with Period V vessels). The best publication of
the modern excavations of Period IV pottery is the analysis and drawings
in Young 1963: 3139, pls. 812; also Young 1965: 55, 7478, figs. 6, 7, and
the photographs in Dyson 1968: 84, 9497. The bridge-spouted vessel, a
characteristic shape from this period, is common in burials, where it was
sometimes (but how often?) placed on a tripod (Ghirshman * 1939: pl. C; Rad
and Hakemi 1950: figs. 20, 23; Dyson 1960c: 121; Crawford 1961: 88, fig. 2). But
many other forms remain unpublished. Again, for this periods pottery, in
burial contexts, see Dinkha Tepe (Muscarella *1968; * 1974; Piller in Stllner,
Slotta, and Vatandoust *2004: 315, Abb. 7).
6 Because of the non-publication of the skeletons recovered in BB IT, I had to acquire the
2004 data from the Hasanlu skeleton (Burial) files (see n. 12 below), which were promptly
supplied to me by Dyson.
320 chapter nine
Concerning Period IV artifacts, in Dyson 1968: 88, we are informed that a

mass of rich objects was recovered; in Dyson 1989b: 110, a great amount; de
Schauensee (1988: 47) says some 2000 metal artifacts were recovered from
this period; Dyson and Voigt (2003: 219) state that more than 1500 artifacts
were recorded from BB II alone; and in Dyson (1959a: 12), 200 artifacts are
cited as deriving from one room of BB I, of which four illustrations are
presented. Dyson 2003: 43 seems to state that 7000 artifacts derived from
Period IV alone; and V. Pigott (in Stllner, Slotta, and Vatandoust * 2004: 252)
states that 2000 iron artifacts, 65 percent being weapons, as well as a similar
amount of bronze artifacts, derived from Period IV. Put succinctly, we do not
know how many artifacts derived from Period IV.
Indeed, a large quantity of artifacts from this period are published and
constitute the major corpus of Iron Age II artifacts known to date-all of
which are associated with the final years of the sites existence, ca. 800bc.
But the very same artifacts are repeatedly mentioned in publications over
the years, sometimes with photographs, and of course including the Gold
Bowl, actually a beaker.7 Across time, one is presented with repeated refer-
ences to helmets, cheek pieces, swords, daggers, spears, axes, knives, maces,
quivers, arrows, metal vessels, jewelry, and so forth, along with many arti-
facts made of iron, silver, bone, gold, stone, etc.; many remain unpublished.
To give but one example: a fragment of an Egyptian Blue vessel was exca-
vated from BB II in 1960 (field no. 60269); in 1964, four other related
Egyptian Blue fragments were excavated from the opposite side of BB II,
7 In an unsigned note, Hasanlu, Iran, in Expedition 1958 (Fall); 35, it is more accurately
described as a large bucketshaped vessel. It was the discovery of the Gold Bowl and the
public attention it received that projected the excavation and its director into prominence.
The first time it was published (spectacularly) was in Life magazine (The Secrets of a Golden
Bowl, January 12, 1959; 5060). References to the vessel by the Hasanlu staff include Dyson
1960c; 1960e; 1960f; 2003; 4546 (in a number of cases the Golden Bowl is included in the
title); also Muscarella 1987, and Winter 1989. Drawn depictions in several publications vary
in small details. It was Charles Burney who made the original drawing at Hasanlu in 1958
under difficult circumstances (a fact mentioned once, in Dyson 1959a; 14); M. de Schauensee
modified it in 1960 from photos, and in 1974 from photos and autopsy in Tehran. The official,
corrected, drawing of the bowl, designated as MTMS 1974 (Dyson 2003: 46) is usually signed
by de Schauensee: see de Schauensee and Dyson 1983: fig. 5; Winter 1989 90, fig. 6; and Marcus
1991: 555, fig. 28. For a discussion of the various drawings and issues of their accuracy, see
Muscarella 1987: 137. That the vessel is in fact a beaker (see the drawing in Winter 1989: 88,
fig. 3) was noted in Muscarella 1989: 34; 1995: 991; and 1996: 211; also by Piller in Stllner, Slotta,
and Vatandoust * 2004: 708. The bowl/beaker was recovered in BB I, Room 9not 6, contra
Dyson 1989b: 124, left; see Winter 1989; 88: fig. 2; on p. 89 Winter says the silver beaker was
recovered in BB I East, but it was BB I West.
and a fifth (field no. 64611a) was recovered in a shop in the nearby town
of Nagadeh. All are clearly related in form, style, and iconography, clearly
reflecting an Assyrian background, either imported from there or closely
adapted (see Porada *1965: 120), and all the fragments derive from one
vessel; in fact, one of the 1964 fragments, the one purchased in Nagadeh,
neatly joins to the 1960 fragment. The 1960 fragment was published in
Crawford 1961: fig. 5; in Dyson 1962: fig. 5; in Porada * 1965: pl. 33; and again
by Dyson in 1966: fig. 16.5, and in 1968: 93but in neither of the latter
two, nor in subsequent reports, were mentions made of the additional
fragments or the join made years earlier. Three of the 1964 fragments were
published for the first time in 2004 in a German museum catalog, courtesy
of the Tehran Museum (Piller in Stllner, Slotta, and Vatandoust * 2004: 733,
no. 405). A number of other Egyptian Blue objects reflecting an Assyrian
background excavated in 1960, 1964, and 1970 also remain unpublished. For
several 1964 published examples of excavated Egyptian Blue objects from
BB II, see Piller in Stllner, Slotta, and Vatandoust * 2004: 711, no. 362, also
710, no. 359 (the same object as Dyson 1972: 4647, fig. 6, and Winter 1980:
2223, fig. 58); another Egyptian object is shown in Winter 1980: fig. 59;
and Muscarella 1966: 133, figs. 30,31, and 33; for the probably Assyrian (or
North Syrian)-made lion bowl, see van Loon 1962 and Muscarella 1965b:
fig. 3.
The corpus of ivories, seals, and glass artifacts is fully published, as are the
imported lion bowls, ivories, and glass. Most of the artifacts are certainly
locally made, called local style, especially among the ivories, seals, lion
pins, and metal productions. For groups of artifacts from various Hasanlu IV
loci published together in one venue, see de Schauensee 1988; and Mus-
carella 1988: 1579. Full or fairly extensive publications exist spread over
many venues:
Ivories: Muscarella 1980.
Glass, imported, including mosaic forms: von Saldern 1966; Marcus
1991; de Schauensee 2001.
Seals: Dyson 1986; Marcus 1989; 1990a; 1990b; 1994a; 1996a (repetitive);
Dyson and Harris 1986.
Metal vessels: de Schauensee 1988; Muscarella 1988: 2432.
Weapons and armor: Hakemi and Rad 1950: pls. 38, 39; Dyson 1964e;
Muscarella 1988: 5363; 1989 (the drawings of alleged Hasanlu warriors
published here in figs. 3 and 10 were added without my knowledge and
are inaccurate. For an accurate, ancient, representation of a Hasanlu
warrior, see Muscarella 1980: nos. 54, 55with earflaps, mustache, and
322 chapter nine
barefoot); Muscarella in Dyson and Muscarella 1989: 1920; Pigott 1989;

de Schauensee 1988 (with no loci given).
Horse equipment: Winter 1980; de Schauensee and Dyson 1983; de
Schauensee 1988; 1989; Dyson 1983/1984: 303; Muscarella 1988: 6572;
Muscarella in Dyson and Muscarella 1989: 1617.
Lion and other pins: Hakemi and Rad 1950: pls. 3133; Marcus 1993;
1994b; 1996b; Muscarella 1988: 4245; and Muscarella 2004.8
Jewelry: Hakemi and Rad 1950: pls. 36, 37, 4245; de Schauensee 1988:
fig. 32; Muscarella 1988: 3241; Marcus * 1995: 24972501.
Imported lion bowls: Muscarella 1965b; 1974; van Loon 1962 (they are
not censors, contra Dyson 1989b: 123124, fig. 22).
Shell artifacts: Reese 1989.
Inscriptions: published in several venues but no corpus exists; none
are local, all are imported.
For discussions of the local Hasanlu style, see Winter 1977: 371375;
1980; Muscarella 1980; Marcus 1996a: 1934, 83102.
Dyson (1989b: 120124) cites 19 categories of selected ornamental items,
some of which are plotted on the plans of BB I and II (but not the other
buildings) to document their specific deposition patterns therein (p. 120,
figs. 18a, b; 19a, b; 23a, b). Only 12 categories of artifacts are plotted in this
manner, and many of them remain unpublished. For example, a number of
glazed objects were recovered in 19601964 from BB II (but not elsewhere;
Dyson 1972: 46; 1989b: 122123, fig. 19), but only a few have been published:
a goblet (Dyson 1968: pl. 39; 1989a: 9, fig. 10b); wall plaques (Dyson 1967:
pl. 1485 B, and another in 1972: 48, fig. 7); and vessels (Muscarella 1966: 132
133, figs. 28, 29). Furthermore, a number of artifacts published over the years
are not accorded a specific locus.
For discussions of foreign imports or local adaptations (aside from im-
ported material), such as ivories, plaques, glass, and lion bowls, see Mus-
carella 1971; 1980: 192199, 200202, 210217; Marcus 1990a; Winter 1977: 375
381. In addition to the Egyptian Blue vessel, there are also a good number
(how many remains unknown) of Assyrian-made (or adapted?) glazed wall
tiles, two of which have been published (Dyson 1959a: 14; 1989a: 9). Dyson
(1989b: 120) reported that from second-floor collapse there were a great
many glazed wall tiles, but he does not distinguish which, or how many,
8 My 2004 article was written to confront claims by Marcus (1993; 1994b; *1995; and 1996b)
about alleged gendered function and cultural use of lion pins and other pins at Hasanlu.
were Assyrian types and which plain glazed tiles (see 1989b: 121, fig. 18b). For
heirlooms, earlier artifacts recovered in IV B, see Dyson and Muscarella 1989:
12, and n. 8.
Relatively speaking, architecture is the most thoroughly published of all
the Hasanlu data, but this information is spread out in various publica-
tions (see figs. 2 and 3). For substantive architecture publications to date,
see Young 1966; 1994; 2002; Dyson 1977b; 1980; 1989b; and Muscarella 1988:
19, n. 4, 208209, n. 3. Several smaller and separate buildings, labeled the
Bead House, South House (directly to the west of BB II), and the Artisans
House, are briefly mentioned. The latter is described as a private house
or a small domestic structure, situated outside the main citadel zone, in
the Outer Town; it contained molds, crucibles, stone vessels, and mortars
(Dyson 1960c: 121122; 1966: 423; de Schauensee 1988: 46; Dyson 1989b: 109);
its locus does not appear on any published plan. The other two buildings are
located close to the northwest corner of BB II. The South House is merely
cited; of the Bead House, it is reported that it contained many beads, bone
handles, and stone bowls, the latter two items being shattered (Dyson 1960c:
122123; Porada * 1965: 112113). None of these structures contents were pub-
lished. The major architecture consists of large, manifestly elite structures,
with multicolumned halls, stairways to an upper level, a roof or second floor
(probably the latter), porticoes, and side storerooms; they are designated as
Burned Building (BB) I, II, and so forth. Throughout the publications, incom-
plete or inconsistent settlement plans are illustrated, creating a source of
confusion. Further, while it is stated in several publications that porticoes
were added to the fronts of several of the Burned Buildings in Period IV B
when they were reconstructed following the IV C fire (BB II, III, and IVIV
E, and V: Young 1966: 55; 2002: 387; Dyson 1965b: 198; 1973a: 303; 1989b: 114;
Muscarella 1996: 210; Dyson and Muscarella 1989: 1), this is explicitly denied
for BB II in Dyson and Voigt 2003: 222, where it is claimed the portico was
present in IV Cbut the earlier claims to the contrary were neither cited nor
explained away.9
9 The caption to fig. 20.1 (Dyson and Voigt 2003: 220, which supercedes a preliminary
[sic] drawing in Dyson 1989b: 112, fig. 6a) is confusing and in contrast to text references about
the portico existing in IV C (p. 222). In a personal communication (February 4, 2005), Dyson
informed me that recent research revealed that BB IIs portico had been added before the
fire terminating IV C, and thus belongs late in that period. However, the portico of BB IV was
added after the fire, constructed in IV B. (Note also that the caption for Dyson 1989b: 117, fig. 13,
an aerial view of the buildings, is completely erroneous.)
324 chapter nine
BB I, built in IV B, is incompletely published in Dyson 1959a: 6, 914,

with a plan; 1960c; 1962; and Young 1966: 5356. The building featured two
rows of irregularly spaced columns. Dyson 1959a is the earliest reference to
artifacts, including the Gold Bowl and silver beaker (see n. 7, above), but
only four artifacts and Assyrian-style wall tiles are illustrated (pp. 1314; the
silver beaker was considered to be of Urartian origin in Dyson 1960f: 250,
fig. 1). Two hundred artifacts are reported as deriving from one room alone;
they include stone bowls and metal vessels, beads, and pottery (p. 12), but
remain essentially unpublished; Winter (1980: 27, figs. 70, 71) published two
bronze belts from here; for others from BB II and a burial, see Muscarella
1988: 4850. The rear, west, and part of the north walls are missing: see the
plans in Dyson 1959a: 6; 1960c: 120; also the aerial photograph in Muscarella
1966: 120; Dyson 1989a: 5, fig. 4; and fig. 3, this article-cf. the restored plan
in fig. 2, this article. Young (1966: 55; 2002: 388) considers BB I to be the
latest Period IV building constructed. Additions and alterations in internal
features are recorded in Dyson 1960c: 123.
BB II has two rows of four columns, plus a ninth column at the north;
it is one of the original buildings built in the IV C settlement (along with
BB III, IV East, and V; Young 1966: 55; 2002: 386). Over time it has become
the most completely published of the periods structures: Dyson 1960c; 1961a;
1961b; 1962 (with a plan); 1963; 1964b; 1965a; 1965b (with a plan: fig. 4); 1968;
1972; and Young 1966. In general, Dyson 1989b and Dyson and Voigt 2003 are
the most complete publications for architectural details to date. However,
the latter publication includes an incompletely drawn plan (fig. 20.1) which,
compared with the earlier Muscarella 1966: 123, fig. 1, is less instructive. BB II
is interpreted by Dyson to have been a religious building or temple (1961a:
535; 1989b: 118119); Piller (in Stllner, Slotta, and Vatandoust * 2004: 325,
n. 17) also posits a cultic function. Young (1966: 5971) rejected a religious
designation, placing BB II within a broad tradition of megaron building
extending from Mycenaean Greece to Bronze Age Anatolia. B. Fehr (* 1971
1972: 4849) cited and accepted Youngs interpretation of BB II and argued
that es sich um den gleichen Grundtypus des reprsentativen Herdhaus
handelt Dandamaev and Lukonin (*1989: 18) argued that all the Hasanlu
structures were intended for civil ceremonies (see also Muscarella 1988:
19, n. 4, for further discussion). A number of its artifacts are mentioned in
many venues, some extensively (ivories, lion pins, seals). In Dyson 1973a:
303; 1989b: 114; and Dyson and Voigt 2003: 222, an earlier destruction and
rebuilding is reported.
BB III, uncovered in 1962, has been briefly cited: Dyson 1964b: 373374;
1965a: 158; 1965b: 198, fig. 5 (with a full plan). The building has only two
central columns, which is unique. Young provides a good summary, suggest-

ing it was a residence (1966: 59); but according to Piller (in Stllner, Slotta,
and Vatandoust *2004: 319), it is a bronze-working center. Nowhere is it men-
tioned that the Period III B fortification wall cut away part of the buildings
rear wall (see Young 1966: 4, fig. 3; and fig. 2, this article; see also below), the
artifacts and their contexts are not published. Although BB III was appar-
ently built in IV C, no IV C fire was mentioned.
BB IV was uncovered in 1970 (Dyson 1971: 170) and described in Dyson
1973c: 3, fig. 7 (a plan). Artifacts, including horse harnesses, are mentioned,
as is one human skeleton outside by the stelae (Dyson 1973c: 3); see also de
Schauensee and Dyson 1983: 6367, and de Schauensee 1989: 40, 51, figs. 6,
27. Dyson and Pigott (1975: 183184) give a brief discussion of BB IV East,
mentioning some artifacts, some with contexts. Dyson 1977b: 551 is also a
brief report, noting that this building was constructed in IV B. Here it is
claimed that the central hall had three pairs of columns, when in fact it has
four central columns (see also BB V, below). Dyson 1983/1984: 302303 also
has a brief description of the building, and numerous skeletons are reported
(p. 302), found in a pile as if they had fallen from an upper levelbut where
were they found? Horse gear from Period IV is published in de Schauensee
and Dyson 1983.
BB V was published for the first time in Dyson 1973a: 304 and 1973c: 12,
fig. 8 (plan).10 A IV C construction and IV B rebuilding are noted. In the 1973c
publication, it is unclear whether the skeletons of four horses discovered
derived from BB IV or V-but Dyson 1973a: 304, and Winter 1980: 31, make it
clear it was in BB V. A fifth horse was also recovered (Winter 1980: fig. 78).
Urine, presumably from the horses, is also reported from the anteroom, and
it is suggested that the building was a stable (Dyson 1973c: 2). In the central
hall four rows of columns (Dyson 1973c: 1) are recorded; however, fig. 8
there shows correctly that the hall in fact has four central columns in two
rows, and nonaligned engaged columns against the walls (fig. 3, this article),
the same as BB IV. The four rows are reported in Dyson 1980: 150with a
plan (see fig. 3; similar to fig. 2, both in this article), which is quite different
from that published in 1973c. The correct number, four central columns,
was given in Dyson 1977b: 551 (with no reference to the earlier error-and
which was repeated in 1980). Some published plans regarding the column
10 Dyson 1973c is a confusing article, as I found it impossible to tell which specific BB
is being described within a section, and there are no figure citations in the text for the
illustrations.
326 chapter nine
positions are misleading in this respect, specifically, Dyson 1997: 479, fig. 1;
Winter 1980: fig. 2; Muscarella 1980: plan 1; Dyson and Muscarella 1989: 2,
fig. 1; and Dyson 2003: 42. For a correctly drawn plan, see Muscarella 1988:
17, fig. 2; 1995: 992, fig. 8; Dyson 1973c: fig. 8; 1989b: 115, fig. 10 (see fig. 3,
this article; but compare in the same journal, Marcus 1989: 58, and Dyson
1989a: 10, fig. 11; see fig. 2, this article; Dandamaev and Lukonin * 1989: 16
is completely wrong). Dyson and Pigott (1975: 183) and de Schauensee and
Dyson (1983) publish brief comments and some artifacts. Young (2002: 386)
says BB V was constructed at the same time as BB II, IV E, and III, but in
1983/1984: 303, Dyson says that V is older than BB II, offering no reference to
the previous claim.
BB IVV, the corridor building between BB V and BB IVE, is mentioned
by Dyson and Pigott (1975: 183), Dyson and de Schauensee (1983: 7175),
and later by Dyson (1983/1984: 303). Many pieces of horse gear/trappings are
listed, including one of the most striking artifacts recovered from Hasanlu:
a bronze breastplate decorated in relief with the figures of a powerful deity
holding two bulls. This magnificent object was fully published by Winter
(1980), along with the juxtaposed horse gear (see also Muscarella 1988: 66
72,7475, nos. 95111, 120).
BB VI and BB VII are the two structures at the west of the mound,
adjacent to the Gateway, and are partially published (Dyson 1975: 181182,
fig. 1; 1989b: 111112, figs. 4, 6b). The Gateway, a major construction with
enclosing walls at the west of the site, was partially published with a plan
in 1975, and interpreted to be an access to the main structures of the citadel.
(This interpretation was challenged by Kroll [* 1992], who argued that the
structure was not a gateway but a stable.)
Replasterings were noticed and counted on a block on the hearth in
BB V. It is claimed that they were burned and re-plastered perhaps 200
times (Dyson 1977b: 551); but three years later the block is now a column,
and the number of replasterings is lowered to over 100 times (Dyson 1980:
150). No reason for the significant adjustment is presented (the latter figure
was chosen by Dandamaev and Lukonin [*1989: 17]).
Many of the contents of all the Burned Buildings remain unpublished
(BB II comes out best here, but often published artifacts from BB II are not
identified as such); no full description of the contents exists for any building
(pace Youngs comment of 40 years ago [1966: 48] about a forthcoming final
report).
Hasanlu IV seems to have had no fortification wall, although in many
publications up to 1975, it was accepted that the citadel was fortified; on this,
see the discussion below for Period III B.
Burials occur in the same cemetery as those of Period V. The number

of Period IV burials is unknown (above), and nowhere have they been
reported meaningfully in any publication; the full contents of not one burial
have been published, Dyson stated (1965a: 158) that sixty-five burials were
recovered from the cemetery, but provided no numeration regarding the
respective cultural periods. Inexplicably (and again unexplained), in the
very same year Dyson (1965b: 209) reported that from Period IV alone there
were nearly 200 graves opened.11 Marcus (*1995: 2499) stated that about
ninety Period IV burials were recovered. And as late as 1989 (Dyson 1989a:
7), we are informed that a number of burials have been recovered in the
cemetery, but no chronological data or specific numbers are provided. Buri-
als are mentioned in Dyson 1958a: 27, with no photos except for their pottery,
but so far as I could determine, photos of only two Period IV burials were
ever published: Crawford 1961: 87, and Dyson 1989b: 109, fig. 3 (the same
as Piller in Stollner, Slotta, and Vatandoust *2004: 314, Abb. 6). In fact, no
record exists concerning how many Period IV burials were uncovered.12 Dan-
damaev and Lukonin (*1989: 20) stated the case succinctly: the burials and
their contents have been published selectively and have been described too
summarily.
Stein published burials from this period along with one photograph and
some contents (1940: 400401, fig. 109, pls. 2425); Ghirshman (* 1939: pl. C)
published drawings of two; see also Hakemi and Rad 1950: figs. 8b, 911, 14,
19, 20, for others. Rathbuns (1972: 1851) Period IV study combines skeletons
from the cemetery and the citadel-but no specific loci are given.
Attempts have been made, based on historical and geographical interpre-
tations, to identify the polity of the Period V and IV populations at Hasanlu,
culturally accepted as being the same across time; and there is much con-
fusion and changing opinions. They have been called Mannaeans (Dyson
1960c: 119120; 1960e: 132; 1961a: 534; 1961b: 64; 1962: 639; 1966: 416; 1967: 2965;
11 Responding to enquiries I made in 1966, Maude de Schauensee informed me there were
100 Period IV burials; in 1987 I was informed by Mary Virginia Harris of the University Museum
that there were 83 from this period. We thus have 200, 100, and 83 Period IV burials!
12 The cause of the disorder results from the recording system in effect throughout the
excavation campaigns: skeletons recovered at the site were registered on a so-called Burial
Sheet and assigned a beta-symbol and number provided in sequence to each skeleton as
encountered. Thus, all skeletons, whether excavated in burials in the cemetery or on the
citadel, or those who perished in the destruction, were all equally registered with a beta
number on Burial Sheets, No separate listing of the burials per se for any period exists (see
n. 11, above).
328 chapter nine
Crawford 1961: 88, 94; Ivantchik *2001: 97, n. 4); Porada (* 1965: 108, 110)
refers to the artifacts as The Art of the Mannaeans. Some have called them
Hurrian speakers (Dyson 1961b: 64; 1962: 642644; 1966: 421); or an unknown,
unidentifiable polity (as in Dyson 1968: 89), where they may be Mannaeans,
but the ethnic identity of periods V and IV is at present inconclusive
(on this, see also Muscarella 1971: 264; 1987: 136). Early on Dyson also
argued for an Indo-European, in fact an incipient penetration of Persians,
at Hasanlu (1963b: 33). Winter (1989: 101103) claimed that there is no solid
evidence to connect Hasanlu V and IV with Indo-Europeans, but allowed for
a Hurrian presence. Young (1985: 368, 374375) also suggested that they were
possibly Indo-European-speaking Iranians.
Salvini, among others, has identified the site as Meshta, a state known
from Urartian texts as existing somewhere in northwest Iran (Salvini * 1995:
25, 4142, 46; Dyson and Muscarella 1989: 19, n. 105). Reade (* 1979) iden-
tifies Hasanlu with Gilzanu (not Meshta, contra Winter 1989: 102), but the
identification essentially hinges on his geographical placements of this and
other polities recorded in Assyrian texts, positions that shift continuously
in modern interpretations; Salvini (*1995: 25) accepts Reades identification
as viable. Also, Reades argument regarding Assyrian material and influence
at Hasanlu, while certain, could be but one example where such manifesta-
tions occurred.
Period III
This period is first mentioned (in Dyson 1959a: 9) as the level recovered
directly above the ruins of Period IV. It was alleged here and in a number
of subsequent publications to be one architectural and cultural unit, often,
until 1999 (below), referred to as the Triangle Ware Phase or period, based
on a painted pottery motif claimed to be present throughout Period III
(Young 1959a: 5; Dyson 1961a: 534; 1962: 641; 1963b: 33; 1967: 2964, 2966,
and fig. 1036). In 1963, Dyson (1963a: 131132) reported that in fact two
distinct levels are present within Period III, the later labeled III A, the
earlier III B (also 1964b: 372; the A, B distinction was ignored in Dyson
1967: 29662967). The recognition of the two levels was not supported or
explained by means of a section drawing (see further below). Young (1965:
5355) saw the two levels as two phases of one cultural period, based on a
contemporary understanding of the painted pottery continuation; but he
did not distinguish between the phases, calling all the pottery (1965: figs. 1,
2) Iron III Thus, the pottery of III B and A was not kept discrete in the
excavation and recording.
The architecture and pottery of these two major cultural periods, III A
and B, still remain sporadically, inadequately, or not published, a condition
noted by Dyson (1999a: 118) as the unpublished state of Hasanlu III data;13
for earlier comments on this, see Muscarella *1973: 71; also Levine * 1987: 234,
who notes the unpublished material from Hasanlu III.
Period III B
Even subsequent to the III B and A stratigraphical distinction, pottery with
triangle decoration continued to be recorded as coming from the III B level
(Dyson 1963a: 132; 1964b: 372; 1965b: 204205, figs. 9, 10, 13; 1966: 419; 1967:
2958, 2960). In Dyson 1972: 46, it is flatly stated that painted wares were the
most identifiable pottery of III B (italics mine); the same was repeated in
1977b: 549 (about which see below). Later, however, in Dyson and Muscarella
1989: 4, the very opposite is claimed-namely, that only a few sherds of
Triangle Ware occur here. Other references to III B include Dyson 1964b
(some discussion); 1965a; 1965b (with more extensive discussion); see also
Dyson 1968; 1977a; 1989a: 56. Aside from Dyson 1964b and 1965b, basically
only pottery is discussed (see discussion below).
The architecture, called Period III over time, is described as simple stone
house foundations (Young 1959b: 65); as rows of rooms (Dyson 1961a:
534); as row houses set side by side (Dyson 1963a: 131); as small rooms
(Dyson 1966: 417); and as a series of single and double barracks-like rooms
(Dyson and Muscarella 1989: 3). But these structures have never been pub-
lished except in a plan labeled Period III, and a fortification wall recorded
as from Period IV, reused in III (below): Dyson 1959a: 9; 1964b: 372, fig. 1. The
period assigned was corrected to III B in the plan published in Dyson 1989a:
7, fig. 6 (fig. 4, this article). Levine (*1987: 234) records (as a personal com-
munication from Dyson) that a re-examination of Period III revealed that
III B had two occupation phases; this was first reported by Dyson much later
(1999a: 132)his section drawing there (1999a: fig. 10) is unclear on this; and
his publication does not mention when or how the phase modification was
recognized.
13 For a remarkable grave accompanied by four horses and arrows (probably socketed)
excavated in 1947 by Rad and Hakerni, see Dyson 1960c: 121; 1965b: 208211; Muscarella 1988:
219220, n. 3; Dyson and Muscarella 1989: 21; Derin and Muscarella *2001: 193; Ivantchik *2001:
185, 276278. It is manifestly post-Period IV B, probably post700 bce, but whether associated
with Period III B or A remains unknown at present.
330 chapter nine
A major stratigraphical adjustment concerning the correct period and

cultural locus of the fortification wall was reported only after the termina-
tion of the excavations. Beginning in Dyson 1959a: 58 (with a plan), and
Young 1959a: 5, and in all subsequent reports until 1975, the fortification wall
was recorded as a Period IV construction (see Dyson 1989a: 56 for explana-
tions) and claimed to have been reused or rebuilt in part in Period III
(Young 1959a: 5; 1959b: 65; 1963: 26, 31, 44; Dyson 1960c: 121; 1961a: 534; 1962:
639649, fig. 2; 1963a: 131132; 1964b: 372, fig. 1; 1965b: 198, 204, fig. 3; 1966: 417,
424; 1967: 29582959, fig. 1028; 1969: 45). Kleiss (* 1973: 85) did note that the
wall looked Urartian, but following the given Period IV attribution, dated it
to the tenth-ninth century bc. E.C. Johnson (* 1976: 25, 37, n. 56: but written
earlier, in 1973) had recognized that the wall in fact overlaid the destroyed
Period IV structure BB III, mentioned above (fig. 2, this article), and conse-
quently she correctly concluded that the wall was constructed in III B and
was Urartian.14
That the fortification wall was a construction of Period III B, not a reuse
from Period IV, was first reported in 1975 (Dyson and Pigott 1975: 185; Dyson
1975: 182); this was reiterated in Dyson 1977b: 549; 1989a: 57, fig. 6 (fig. 4,
this article); and in Dyson and Muscarella 1989: 34, fig. 2, and pp. 20, 22.
In the first cited publication (Dyson and Piggott 1975: 185), it was reported
that the 1974 campaign demonstrates conclusively through stratigraphical
evidence that the entire wall belongs to the III B period and not just a
section of it as previously hypothesized [italics mine]. Dyson added that
in 1972 Kroll had pointed out to me that the wall reflects the architectural
practices [italics mine] of Urartian sites. In the second 1975 publication
(Dyson 1975: 182), Dyson states that excavations in 1972 and 1974 now prove
conclusively what has been suspected for some timebut how much time,
and which suspicions existed, had never previously been mentionedthat
the wall belonged to III B. He added our own stratigraphic investigations
along with Dr. Krolls arguments on ceramic [italics added] grounds, and
Dr. Kleisss analysis of Urartian architectural history confirm one another.
In the third (Dyson 1977b), it was Krolls pottery parallels of the III B pottery
with Bastam, and suggestions on architectural grounds by W. Kleiss (both
14 This article presented many good points. But parts of the article are confused. Johnsons
n. 45 (information given to her by Muscarella) states that the wall was indeed built in
Period III, not IV, but this note is a later addition to the manuscript shown to me; see
my comments on this issue in Muscarella 1980: 214, n. 39, and in Dyson and Muscarella
1989: 23 n. 11. It is possible that the alleged stratigraphy of the fortification wall lay behind
R.D. Barnetts erroneous claim (* 1982: 321, 341) that Period IV was an Urartian levelbut
his n. 60 is meaningless as a reference.
Figure 4. Plan of Hasanlu III B (courtesy of the Hasanlu Project).
given verbally), that led in 1974 (1972 was not cited here) to a reexamination
to test the hypothesis that the wall was constructed in III B. Finally, in the
fourth reference (Dyson 1989a), there is no mention of the background that
led to recognition of the accurate construction date. Rather, it is presented
as a normal ongoing result of excavations in 1972; included is a photograph
(1989a: fig. 7) of the trench that clarified the stratigraphy. But there is no
mention of the contributions of Kroll or Kleiss.
Also published in 1989 is a photograph taken in 1962 (Dyson 1989a: 5,
fig. 4; see also Muscarella 1966: 120) that clearly shows the III B wall cutting
into and destroying BB Is rear and side walls-very clearly shown in the
plan published in Dyson 1959a: 6. Young (1966: 53) had concluded that
foundation trenches dug for the structures of Period III had destroyed
the rear wall of BB I. But the implication of these observations was missed,
because it was a given then that the wall was a Period IV construction.
332 chapter nine
Although this is based on personal recollections and cannot be con-

firmed, in 1972 Kroll allegedly indicated that the stratigraphic and construc-
tion evidence showed that the foundation trench for the fortification wall
cut into Period IV, that it was indeed from Period III B, post-Period IV, Urar-
tian (the pottery was not at issue here). Further, during the following cam-
paign in 1974, Kroll and Kleiss and others were said to have reexamined the
wall and foundation trench and agreed that it was indeed a post-Period IV
construction. Paradoxically, and notwithstanding the 1975 and 1977 redat-
ing, the III B fortification wall continued to be published in a plan together
with the Period IV settlement (Dyson 1977a: 157, fig. 1; 1989a: 10, fig. 11; 1997:
479, fig. 1 [see fig. 2, this article]; 2003: 42), with a note that the wall belonged
to Period III; cf. Dyson 1989a: 7, fig. 6, where the wall is correctly depicted in
a Period III B plan (fig. 4, this article).
As noted above, pottery attributed to Period III B is mentioned in many
publications, sometimes with drawings, beginning in Dyson 1959a: 9, de-
scribed as decorated with hanging triangles similar to sherds recovered by
survey at Ziwiye in 1957. That Triangle Ware present in III B continued into
III A is recorded many times: Dyson 1960e: 132; 1961a: 534; 1963: 132; 1964b:
372; 1977b: 549. In Dyson 1965b: 204, it was suggested that it may have been
imported; and on pp. 205207, parallels with Sialk and Ziwiye are cited, and
fig. 9 illustrates Hasanlu III B and Ziwiye sherds; fig. 10 depicts Hasanlu
IIIB Triangle Ware vessels. Young (1965: 55) records that III B and A both
have painted pottery, and his fig. 2 is labeled Hasanlu III-see also fig. 5; see
also Dyson 1969: 45. Young (1965: fig. 5) published this same Ziwiye-Hasanlu
III B drawing, for which also see Dyson 1967: 2964; 1972: 46. Compounding
the confusion, Levine (*1987: 234) reported (quoting Dyson) that painted
wares derive only from the upper III B phase and the later III A contexts.15
At the time of early reporting on Period III, a well-polished red vessel
was cited (Dyson 1959a: 9) as paralleled at Toprakkale of the eighth century
bc (a century too early in date), but no photo was provided. And Young
(1963: 2829) also described a red polished, burnished ware, but did not
recognize it as Urartian pottery. In 1964b: 372, Dyson reports that alongside
painted pottery in III B, [t]he ubiquitous burnished red-slipped bowls of
the period seem to have Urartian parallels, for which he gives a photograph
(p. 364, fig. 7but mislabeled as from Period IV).
15 See also Kroll * 1976: 164. Such was the confusion that P. Calmeyers entry Hasanlu in
the Reallexikon der Assyriologie (Calmeyer 1975: 129) states that III A contained only plain
ware and III B Piitnted Triangle Ware.
Also published in 1964, recovered on the citadel, is a fibula found inside

the bowl (Dyson 1964b: 364, fig. 5) that indicates Urartian influence; it is
in fact an Urartian fibula (see below). We thus learn here that many typ-
ical Urartian vessels and a typical Urartian artifact were recovered from
Period III B, but their implications-that the settlement was Urartian-were
not then recognized. When I published the same vessel and fibula (Mus-
carella 1965a: 237, figs. 1, 2), I attributed them to III B and cited the Urartian
parallels, but given the prevailing perception I avoided calling III B an Urar-
tian settlement (for another Urartian fibula, see below).
S. Kroll was the first scholar to argue that it was only angeblich that III
B had Triangle Ware, and that this ware indeed does not exist in the Urar-
tian period, but developed later (*1975: 7374; * 1976: 164165; * 1977: 105, and
n. 117), Dyson addresses this issue, one that now indeed reflects the reality
at Hasanlu, only decades later (1999a: 118, 134). It was asserted that after a
careful reanalysis of the field records he determined that the characteris-
tic pottery labeled Triangle Ware erroneously ascribed to III B was in fact
confined to III A, the post-Urartian level. But Krolls earlier important and
correct observations were not cited, nor was it recorded how the earlier erro-
neous assertions had come to exist in the first place. Triangle Ware was here
divided into two stylistic categories: Classic, claimed to be imported; and
coarse, considered to be locally made (repeated in Dyson 1999b). In Dyson
1999a: fig. 12, three vessels from III B are illustrated; they are juxtaposed with
a triangle-decorated pithos and a doublehandled vessel, correctly reported
here as from the nearby Urartian site of Agrab Tepe (Muscarella * 1973: 57,
figs. 15: 16, figs. 23, 24), in Hasanlu terms Period III B. In two previous publi-
cations (Dyson 1965b: fig. 13; 1968: 84), however, the two Agrab Tepe vessels
had been erroneously published as from Hasanlu Period III A. The 1999a,
fig. 12 label merely notes Revised Pottery Summary. Replacing Dyson 1965:
fig. 13 (1968 is not mentioned). There is no explanation given for the change,
or what had caused the earlier significant error (see n. 3).
The stratigraphy of Period III B and A was discussed for the first time
in Dyson 1989a. Dyson reported (pp. 135, 137) that Triangle Ware pottery
appears over a hard and well-formed erosion surface, and a section draw-
ing (fig. 10) is published here for the first time (p. 132). It depicts two over-
lapping units, with the trench labels Operation BB28 and CC28 and DD28
(pace the caption, the section does not include Operations Z28 and AA28
and only includes a small part of BB28); the year this section was drawn
is not mentioned. Descriptions of Period III stratigraphy presented in the
text are unclear. Indeed, the stratigraphical separation of III B and what
exists above seems to be clear in the fig. 10 section: but this level is labeled
334 chapter nine
Period II/IIIA. And the erosion surface was not recorded in any previous
publication, although, as noted above, he had earlier (in 1963a: 131132 and
1964b: 372) called attention to the two, A and B, phases. The visual mes-
sage of fig. 10 also does not correlate with the section published years earlier
(see below), nor does it relate to the descriptions of the stratigraphy given
in Young 1965: 55 (and also in Dyson 1965b: 212) that the shift from the
occupation of III B to III A is not clear due to erosion, and that some III
B walls continued in use or served as foundations for new walls of III A.
These 1965 statements are in conflict with the description and the section
published 35 years later. If the newly published section and data accurately
reflect the excavated reality, the clear separation of III B from the succeed-
ing level, how could the excavator have asserted vigorously for decades that
III B contained painted wares-which were in fact not present there?
Only one other section drawing for the Hasanlu excavations had previ-
ously been published (probably drawn in 1960). It appeared three times
(Dyson 1962: 640; 1966: fig. 16-3; 1967: 2962, fig. 1032), with a thin layer labeled
III. Albeit not stated in Dyson 1999a (it took me some time to recognize
this), this earlier published section depicts the area directly adjacent to that
shown in 1999a: fig. 10. Note that the site locus trench labels given, Op.
XLII, Op. XXIX, were later changed to AA28 and BB28 (none of these trench
identifications were ever shown in any published plan). We are not given
expected information either about the locus on the site for the separately
published, contiguous sections, their position on the published plan of III
B (see fig. 4, this article), or why the Period III remains are labeled differ-
ently in the earlier (III) and later published sections (II/IIIA). I was able
to determine that the two sections depict the area immediately adjacent to
the south wall of BB II, extending to the fortification wall of III B. In this area,
only one isolated Period III B building was excavated (fig. 4, this article). Why
was not a section published of the stratigraphical sequences of Periods III B,
III A, and IV in the only areas on the site where Period III remains are con-
centrated, adjacent to the fortification wall, and here apparently above BB I
and III? (A number of section drawings were made over the years.)
Of the artifacts recovered, few are published. These include bone psalia
(Dyson 1964b: 372, figs. 2, 3): the former figure is labeled as from Period III,
and the latter is mislabeled as from IV in the caption, but from III B in the
text; in Dyson 1965b: 211, it is also assigned to IV, here with a major but incor-
rect historical conclusion regarding the presence of Scythians in northwest
Iran in the ninth century bc. This error (never corrected by Dyson) was inad-
vertently used by several scholars to arrive at the same wrong conclusion
about a Scythian presence outside their homeland (corrected in Muscarella
*1974: 79, n. 16; 1988: 66, n. 2, and 220, n. 3; Derin and Muscarella * 2001: 199
and n. 83; see also Ivantchik *2001: 277, n. 66).16 In Dyson 1964b there is also
the fibula mentioned above; and another Urartian fibula was recovered in a
(still unpublished) tomb (Muscarella 1966: 135, fig. 37; 1988: 4748, no. 53),
which I labeled III A or B; it is surely from III B. In Dyson and Muscarella
1989: 20, eight Urartian fibulae are cataloged as from III B. Dyson (1965b: 207)
also mentions a socketed trilobate (?) arrow, recovered from fill dumped
from the III B structures; a socketed bilobate arrow was incorrectly pub-
lished by me as from Period IV (Muscarella 1988: 63, no. 90, but see p. 107)
in fact, it derived from Period III B (Derin and Muscarella * 2001: 193, and
n. 16). Two Urartian-style stamp seals were published by Marcus (1996a: 147
149, figs. 117, 118) as deriving from two tombs outside the fortification walls.
Neither tomb is published, and hence other contents remain unknown-an
unfortunate omission inasmuch as (unplundered) Urartian burials are rare,
at Hasanlu and elsewhere.
For Period III B burials, we have no information other than that they were
excavated (above). Dyson (1965a: 158) refers to a number of Period III burials
found in the upper layer of the Citadel (see Period II, below), but which
locations continue to remain unknown; nothing is mentioned here about
cemetery burials. Rathbun (1972: 1215) refers to 13 skeletons from this period
that he examined, but he does not mention whether they came from the
cemetery or the citadel.
Period III B was dated in Dyson 1965b: 204, to 700600 bc; in 1973c: 1, to
650600bc; in 1972: 46, to between (?) 750600 bc; in Dyson and Pigott
1975: 182, to ca. 750600 bc; in Dyson 1989b: 110, to 8th and 7th century bc;
and in Dyson and Muscarella 1989: 8, 19, 2021, to between 800 and 600bc,
and probably 7th century. The issue of its chronological relationship to the
earlier IV B destruction and IV C squatters phase was not discussed except
by Young (1963: 4748), who ponders that a long gap may have existed.
The Dyson 1973c publication was the first to mention that Period III B
was an Urartian settlement (except for the fortification wall), which can
be considered a historical reality (Dyson and Muscarella 1989: 4, 19, 20;
Muscarella 1989: 3435; Dyson 1989b: 104). The terminal date of ca. 600bc
+/- is approximately correct: III B (and Agrab Tepe) was probably destroyed
at the same time as many other Urartian citadels, during the reign of Rusa II,
or possibly later. But precisely when III B was built after 800 bc remains
unresolved.
16 See also Young * 1967: 26, 33.

336 chapter nine
Agrab Tepe, another Urartian site (a fort) close to Hasanlu, had been
destroyed twice in its history (Muscarella * 1973). Partly because of the pub-
lished reports that both Hasanlu III B and III A contained painted and plain
wares (Muscarella *1973: 65, 71, 74), I did not then recognize that the com-
plete absence of painted wares at Agrab and the presence of manifest Urar-
tian pottery there, as well as an Urartian sealing (Muscarella * 1973: 6265;
see also pp. 69, 7374), along with the fortification wall similarity, might indi-
cate that Agrab was Urartian, and contemporary solely with the III B period
at Hasanlu (Dyson and Muscarella 1989: 4, 19, 21). It was indeed an idea I
had considered, along with the possibility that the site may have contin-
ued into the post-III B period (Muscarella * 1973: 69, 74), but I could draw no
conclusions. The confusion is now obvious: I had used Hasanlu alleged III
A pottery (viz. as noted above, the pithos with triangles: a classic Urartian-
III B-vessel!) along with III B pottery parallels. In 1989 (Muscarella in Dyson
and Muscarella: 19, 21), I discussed Hasanlu III Bs Urartian history and its
incipient and terminal chronological problems.
A major omission in the Hasanlu publications regarding III B warrants
special mention, namely that, as recognized by P. Zimansky, there has been
a lack of recognition and hence relevant discussion of the Urartians as a
significant presence there.17 This omission occurs as well elsewhere in the
literature (viz. in Salvini *1995). I suggest that this is because the III B and
A division has not been articulated in the Hasanlu publications, nor has the
nature of the complete cultural differences between the two important and
distinct cultural periods, III B and A, been emphasized.
Period III A
No architecture is published for Period III A, and it is reported only (Dyson
1963a: 132) that this level is less impressive [than III B], with more scattered
structures. In Dyson 1964b: 372, it was claimed that the unique painted ves-
sel, recovered fragmented and out of context (see Dyson 1961a: 536, fig. 8),
is now known to belong to III A; also in 1965b: 212. However, a few years
later (Dyson 1967: 2960, fig. 1030),18 the vessel was dated to either Period IV or
17 P. Zimansky (* 1995: 112, n. 14). The Urartian sites of Agrab Tepe and Qalatgah (see n. 19
below) suffer the same fate: Zimansky ignores them in his discussion of Urartian sites in
northwest Iran. Based on the published record, he credited Dyson for recognizing that the
fortification wall was built in III B; he also oddly claims that no Urartian pottery was recorded
there.
18 Grace Freed Muscarella made the drawing, although this was not noted in the caption.
And the bone furniture tripod unit-possibly a drum-from BB II illustrated in Dyson and Voigt
III, but without any reason given for the hesitancy and change (stratigraphy
confusion?). That uncertainty exists regarding III A pottery is exemplified by
the pithos and double-handled vessel (from Agrab Tepe) published as from
Hasanlu III A and III B in Dyson 1965b and 1999a (mentioned above), the
uncertainty is unrelieved by the fact that Young (1963: 30) said that painted
pottery continues into III A, while in 1972: 47, Dyson claimed that there is
no painted pottery in III A, and yet in 1977b: 549, the contrary was correctly
reported, but with no reference to or explanation of earlier information
provided. As observed above, this situation changed dramatically in Dyson
1999a, which states that III A alone yielded painted wares. In 1999b, Dyson
repeats much of 1999a (they are essentially one article split into two); but
1999b has more drawings of the painted pottery, Not mentioned in Dyson
1999a or 1999b is the fact that at the Urartian site of Qalatgah (Muscarella
*1971), ca. 15 miles west of Hasanlu, painted III A sherds were recovered,
demonstrating also that sites post-Urartian history.19
Regarding chronology, we have shifts throughout the publication record,
as follows. Dyson 1965b: 212, gives 600400bc; in 1972: 47, a provisional date
of 400 bc is suggested; Dyson and Pigott 1975: 182, 185, claim 600-?-300bc,
Achaemenian period or shortly thereafter; Dyson 1977b: 549, alleges dated
to Achaemenian times sometime after 600bc; likewise in 1999a: 137; and
Dyson in Dyson and Muscarella 1989: 8, suggests around 400bc In 1999a:
135, Dyson posits that III A commenced long after the abandonment of
III B, perhaps a century later, but correctly states that we do not know the
absolute date.
Period II
In Dyson 1959a: 9, and Young 1959a: 4, we are told that stone-lined tombs
(how many?) recovered in the debris of Period III could belong to Period II.
In Dyson 1957: 39; 1959a: 9; and 1961a: 534, there is reference to cist tombs that
could be from the Parthian period, or Period II; no photos were published.
In the latter publication-and thereafter-II is referred to as the Mystery
2003: 231, fig. 20.7, was both excavated and drawn by me, rather than drawn by the individual
credited in the caption (he was not at Hasanlu). Compare the credit information given in
the same article, the caption on p. 220, fig. 20.1; see also Dyson 1989b: 115, fig. 10. The plan for
BB II was drawn by its various excavators, assisted in 1964 by Edward Keall, then serving as
architect. Each excavator drew a plan of the area s/he excavated in the field, which was then
joined to a master plan; the latter was thus enlarged and modified over the years. The plans
of BB I were drawn by both T. Cuyler Young and Dyson, and that of BB III by its excavator
T. Cuyler Young. (I do not know who drew the plans for the other buildings.)
19 For an account of the discovery of Qalatgah on July 31, 1968 (not July 30, as per
338 chapter nine
Period, consisting of a very large building, set on a platform. In Dyson 1962:

639, stone cist tombs are again mentioned without the quantity given. In
Dyson 1967: 2966, the structure is described as a massive set of foundations;
in Dyson 1972: 5152, as a large building which covered an area more than
24 meters square. Also mentioned here as from this period (Dyson 1972:
51) are two stone-lined tombs containing double burials accompanied by
fibulae. In one of these tombs, one skeleton wore a typical Near Easterntype
fibula, while the other had a hinge-which suggests an Urartian (type) fibula
(I have no photo or drawing of it; for Urartian fibulae, see above). The former
is described in Muscarella 1966: 135, fig. 38 and 1988: 4546, no. 52 (here with
II? given for its level). This tomb was recovered in the burnt fill of BB II.
I believe the level from which it was dug remains to be investigated: is it a
Period II burial?
Rathbun (1972: 1011) reports on three skeletons from this period, two
(although he does not mention this) apparently from one of the two cist
tombs mentioned above. Thus we learn indirectly that there are at least
two cist tombs from Period II. In Dyson 1977b: 548, it is revealed that the
still not illustrated structure has large square and rectangular rooms and a
stairway and hearth. Decades later (Dyson 1999a: 132; 1999b: 101), a bit more
information is squeezed out: the structure consists of subterranean rooms
around two sides of an open courtyard, and there is a fortification wall (Wall
2). Neither a full description nor a plan of this major structure has ever been
published.
As for chronology, in Dyson 1957: 39, and 1959a: 9, Period II is called
Parthian; 1961a: 534, Achaemenian to Sasanian; in 1967: 2966, and 1973b: 195,
it is late Achaemenian; 1968: 85, after 400 bc; in 1972: 51, it is difficult to
Muscarella * 1971a: 45), when I was directing the excavation of nearby S Girdan, see Mus-
carella * 1969: 56, n. 4, and * 1971a. A report on the 1968 campaigns at Haji Firuz, Dinkha
Tepe, S Girdan, and Qalatgah was published by Dyson with the names Muscarella and Mary
Voigt added (* 1969: 179181). Voigt was director at Haji Firuz; I the director at S Girdan, who
on an exploration trip discovered Qalatgah; and Dyson was director of Bronze Age levels at
Dinkha Tepe. Another brief report is in Dyson * 1969: 19. After the discovery and several sur-
veys of Qalatgah, I invited the Dinkha team to the site; they came for the first time on August
9. My encounter and surveys of the site immediately revealed fortification walls (thought to
be a road by the locals), which, together with the surface finds of pottery, manifested the Urar-
tian nature of the site. On the survey with the Dinkha staff, an Urartian seal and an Urartian
stone inscription were discovered, the latter by a member of the S Girdan team, Christopher
Hamlin. For the inscription, see Muscarella * 1971a: 4748, and M. van Loon (*1975).
Dysons brief statements in 1969: 44, Dyson * 1969: 19, and in Dyson, Muscarella, and Voigt
1969: 181, concerning Qalatgah do not reproduce the true nature of the events. However,
M. van Loon 1975: 201, n. * got it right, S. Kroll (in Stllner, Slotta, and Vatandoust *2004: 363)
suggests that Qalatgah was the predominate military fortress in the area, not Hasanlu III B.
date; in 1973c: 1, it is pre-Parthian; in 1999b: 101, 105, Hellenistic pottery and

a Greek sherd (error?) are mentioned but not illustrated, dated pre-275bc.
Period I
The top level at Hasanlu, the Islamic Period, has been published by Danti
2004.
Conclusions
As a conclusion, I refer the reader back to the introduction above, the verac-
ity of which indictments I believe my text has, alas, amply demonstrated.
I add a couple of related comments: [F]ailure to produce a full published
excavation report is simply tantamount to wanton destruction of an archae-
ological site. [T]he excavation of a site is never warranted unless com-
prehensive publication is undertaken, nor is the excavation complete until
publication is ended (Renfrew *1980: 295). After this paper was com-
pleted, I came across Mousavi *2005. In pithy language and understandable
frustration, he writes of Hasanlu: The final report of the excavations has
never been published, and the preliminary reports are either in the form
of news or general syntheses (2005: 92 n. 11)this in 2005, 50 years after
the beginning of excavations. Renfrew and Mousavi both communicate suc-
cinctly one of the failures of the Hasanlu publications analyzed above.
Acknowledgments
I wish to thank the two colleagues who carefully read a draft of this paper
or who communicated with me via email. They gave me what I expected
valuable comments, criticisms, and suggestions, all of which helped me very
much. I, of course, am responsible for what advice I did not take.
Addendum:
Reports or Brief Mention of 19561964,
19701974 Hasanlu Campaigns
1. 1956: Dyson 1957

2. 1957: Dyson 1958a; 1958b
3. 1958: Dyson 1959a; 1960e; 1960f; Young 1959a; 1959b
4. 1959: Dyson 1960c
340 chapter nine
5.1960: Dyson 1960d; 1961a; 1961b; Crawford 1961

6.1962: Dyson 1963a; 1963b; 1964b
7.1964: Dyson 1965a; Muscarella 1966
8.1970: Dyson 1971; 1973a; 1983/1984: 302303. (Dyson 1999a: 139 cites
Excavations at Hasanlu in 1970 as in press for A Survey of Persian
Art 18: 33573370. That publication had not come out as of May 2006.)
9. 1972: Dyson 1973a; 1973b; 1973c
10. 1974: Dyson 1975; Dyson and Pigott 1975
Chronological List of References on Hasanlu
1940
Stein, A.: Old Routes of Western Iran. London: Macmillan.
1950
Hakemi, A., and Rad, M.: The Description and Results of the Scientific Excavations at
Hasanlu, Solduz. (Trans. P. Barzin, from Persian). Guzarishha-y Bastan Shinasi,
1329 A.H.: 87103. Tehran: Vizarat-i Farhang, Idarah-i Kull-i Bastanshinasi.
1957
Dyson, R.H., Jr.: Iran: 1956. University Museum Bulletin 21-1: 2539.
1958
Dyson, R.H., Jr.:
1958a: Iran 1957: Iron Age Hasanlu. University Museum Bulletin 22/2: 2532.
1958b: Pennsylvania Campaign in Iran. Archaeology 11: 128.
1959
Dyson, R.H., Jr.:
1959a: Digging in Iran: Hasanlu, 1958. Expedition 1/3: 417.
1959b: The Silver Cup of Hasanlu. Archaeology 12: 171.
Young, T.C., Jr.:

1959a: Excavations at Hasanlu: The 1958 Season. Bulletin of the Philadelphia
Anthropological Society 12/2: 47.
1959b: Successful Season at Hasanlu. Archaeology 12: 6566.
1960
Dyson, R.H., Jr.:
1960a: The Death of a City. Expedition 2/3: 211.
1960b: Hasanlu-Azerbaijan Project. Explorers Journal 38/4: 1011.
1960c: Hasanlu and Early Iran. Archaeology 13: 118129.
1960d: Expedition News, Hasanlu, Iran. Expedition 3/1: 11.
1960e: Where the Golden Bowl of Hasanlu Was Found: Excavations near Lake
Urmia Which Throw New Light on the Little-Known MannaeansPart 1.
Illustrated London News 236, January 23: 132134.
1960f: The Golden Bowl and the Silver Cup-Treasures with a Dramatic History
and a Rich Significance: Excavations at Hasanlu, near Lake UrmiaPart II.
Illustrated London News 236, February 13: 250251.
1961
Crawford, V.E.: Hasanlu 1960. Bulletin of the Metropolitan Museum of Art 20: 8594.
Dyson, R.H., Jr.:
1961a: Excavating the Mannaean Citadel of Hasanlu; and New Light on Several
Millennia of Persian Azerbaijan. Illustrated London News 239, September 30:
534537.
1961b: Hasanlu, 1960 Campaign. Archaeology 14: 6364.
1962
Dyson, R.H., Jr.: The Hasanlu Project. Science 135/3504: 637647.
van Loon, M.N.: A Lion Bowl from Hasanlu. Expedition 4/4: 1419.
1963
Dyson, R.H., Jr.:
1963a: Hasanlu Discoveries, 1962. Archaeology 16: 131133.
1963b: Expedition News. Expedition 5/2: 33.
Young, T.C., Jr.: Proto-Historic Western Iran. An Archaeological and Historical Re-
view: Problems and Possible Interpretations. Ph.D. dissertation, University of
Pennsylvania.
1964
Dyson, R.H., Jr.:
1964a: A Stranger from the East. Expedition 7/1: 3233.
1964b: In the City of the Golden Bowl: New Excavations at Hasanlu in Persian
Azerbaijan. Illustrated London News 245, September 12: 372374.
1964c: Ninth Century Men in Western Iran. Archaeology 17/1: 311.
1964d: Sciences Meet in Ancient Hasanlu. Natural History 73/8: 1625.
1964e: Notes on Weapons and Chronology in Northern Iran around 1000BC, pp.
3245 in Dark Ages and Nomads c. 1000bc: Studies in Iranian and Anatolian
Archaeology, ed. M. Mellink. Istanbul: Nederlands Historisch-Archaeolo-
gisch Instituut.
1965
Dyson, R.H., Jr.:
1965a: Hasanlu Excavations, 1964. Archaeology 18: 157159.
1965b: Problems of Protohistoric Iran as Seen from Hasanlu. Journal of Near East-
ern Studies 24: 193217.
342 chapter nine
Muscarella, O.W.:
1965a: A Fibula from Hasanlu. American Journal of Archaeology 69: 233240.
1965b: Lion Bowls from Hasanlu. Archaeology 18: 4146.
Young, T.C., Jr.: A Comparative Ceramic Chronology for Western Iran, 1500500bc
Iran 3: 5385.
1966
Dyson, R.H., Jr.: The Hasanlu Project. Pp. 413435 in New Roads to Yesterday, ed.
J.R. Caldwell. New York: Basic.
Muscarella, O.W.: Hasanlu 1964. Bulletin of the Metropolitan Museum of Art 25: 120
135.
von Saldern, A.: Mosaic Glass from Hasanlu, Marlik, and Tell al-Rimah. Journal of
Glass Studies 8: 925.
Young, T.C., Jr.: Thoughts on the Architecture of Hasanlu IV. Iranica Antiqua 6: 4871.
1967
Dyson, R.H., Jr.: Early Cultures of Solduz, Azerbaijan. Pp. 29512970 in A Survey of
Persian Art 14, ed. A.U. Pope. London: Oxford University.
1968
Dyson, R.H., Jr.: Hasanlu and the Solduz and Ushnu Valleys: Twelve Years of Explo-
ration. Archaeologia Viva 1: 83101.
1969
Dyson, R.H., Jr.: A Decade in Iran. Expedition 11/2: 3947.
1971
Dyson, R.H., Jr.: Hasanlu. Iran 9: 170.
Muscarella, O.W: Hasanlu in the Ninth Century bc and Its Relations with Other
Cultural Centers of the Near East. American Journal of Archaeology 75: 263266.
1972
Dyson R.H., Jr.: The Hasanlu Project, 19611967. Pp. 3958 in The Memorial Volume
of the Vth International Congress of Iranian Art and Archaeology, Tehran-Isfahan-
Shiraz, 11th18th April 1968, Vol. 1. Tehran: Ministry of Culture and Arts.
Rathbun, T.A.: A Study of the Physical Characteristics of the Ancient Inhabitants of
Hasanlu, Iran, Miami: Field Research Projects.
1973
Dyson R.H., Jr.:
1973a: Further Excavations at Tepe Hasanlu, Iran. Archaeology 26: 303304.
1973b: Hasanlu. Iran 11: 195196.
1973c: Hasanlu 1972. Proceedings of the 1st Annual Symposium on Archaeological
Research in Iran, ed. F. Bagherzadeh. Teheran: Iranian Centre for Archaeo-
logical Research (paper published as separate fascicle).
1974
Muscarella, O.W.: The Third Lion Bowl from Hasanlu. Expedition 16/2: 2529.
1975
Dyson, R.H., Jr.: Hasanlu 1974: The Ninth Century BC Gateway. Pp. 179188 in Proceed-
ings of the IIIrd Annual Symposium on Archaeological Research in Iran, 2nd7th
November 1974. Tehran: Iranian Centre for Archaeological Research.
Dyson, R.H., Jr., and Pigott, V.C.: Hasanlu. Iran 13: 182185.
1977
Dyson R.H., Jr.:
1977a: Architecture of the Iron I Period at Hasanlu in Western Iran, and Its Impli-
cations for Theories of Migration on the Iranian Plateau. Pp. 155169 in Le
plateau iranien et lAsie centrale des origines la conquete islamique, ed.
J. Deshayes. Paris: ditions du Cente national de la recherche scientifique.
1977b: The Architecture of Hasanlu: Periods I to IV. American Journal of Archaeol-
ogy 81: 548552.
Winter, I.: Perspective on the Local Style of Hasanlu IVB: A Study in Receptivity.
Pp. 371386 in Mountains and Lowlands: Essays in the Archaeology of Greater
Mesopotamia, eds. L.D. Levine and T.C. Young, Jr. Bibliotheca Mesopotamica 7,
Malibu: Undena.
1980
Dyson, R.H., Jr.: The Question of Balconies at Hasanlu. Pp. 149157 in From Athens to
Gordion: The Papers of a Memorial Symposium for Rodney S. Young, ed. K. DeVries.
University Museum Papers 1. Philadelphia: University Museum, University of
Pennsylvania.
Muscarella, 0. W.: The Catalogue of Ivories from Hasanlu, Iran. University Museum
Monograph 40; Hasanlu Special Studies 2. Philadelphia: University Museum,
University of Pennsylvania.
Winter, I.J.: A Decorated Breastplate from Hasanlu, Iran. University Museum Mono-
graph 39; Hasanlu. Special Studies 1. Philadelphia: University Museum, Univer-
sity of Pennsylvania.
1983
de Schauensee, M., and Dyson, R.H., Jr.: Hasanlu Horse Trappings and Assyrian
Reliefs, Pp. 5977 in Essays on Near Eastern Art and Archaeology in Honor of
Charles Kyrle Wilkinson, eds. P. 0. Harper and H. Pitmann. New York: Metropoli-
tan Museum of Art.
1983/1984
Dyson, R.H., Jr.: Summary of Work at Hasanlu and Hissar, 19701980. Archiv fr
Orientforschung 2930: 302304.
344 chapter nine
1985
Young, T.C., Jr.: Early Iron Age Iran Revisited: Preliminary Suggestions for the Re-
analysis of Old Constructs. Pp. 361378 in De lIndus aux Balkans: Recueil la
mmoire de Jean Deshayes, eds. J.-L. Huot, M. Yon, and Y. Calvet. Paris: Recherche
sur les Civilisations.
1986
Dyson, R.H., Jr. and Harris, M.V.: The Archaeological Context of Cylinder Seals
Excavated on the Iranian Plateau. Pp. 79110 in Insight through Images: Studies
in Honor of Edith Porada, ed. M. Kelly-Buccellati. Bibliotheca Mesopotamica 21.
Malibu: Undena.
1987
Levine, L.D.: The Iron Age. Pp. 229250 in The Archaeology of Western Iran: Settle-
ment and Society from Prehistory to the Islamic Conquest, ed. F. Hole. Washington,
DC: Smithsonian Institution.
Muscarella, O. W, Review of Problmes concernant les Hurrites, Vol. 2, ed. M.-T. Bar-
relet. Journal of the American Oriental Society 107: 135137.
1988
de Schauensee, M.: Northwest Iran as a Bronze-Working Centre: The View from
Hasanlu. Pp. 4562 in Bronzeworking Centres of Western Asia c. 1000539bc, ed.
J.E. Curtis. London: Kegan Paul.
Muscarella, 0. W.: Northwest Iran, Excavated Objects, The Hasanlu Project. Pp. 1581
in Bronze and Iron: Ancient Near Eastern Artifacts in The Metropolitan Museum of
Art, by 0. W. Muscarella. New York: Metropolitan Museum of Art.
1989
de Schauensee, M.: Horse Gear from Hasanlu. Expedition 31/23: 3752.
Dyson, R.H., Jr.:
1989a: Rediscovering Hasanlu. Expedition 31/23: 311.
1989b: The Iron Age Architecture at Hasanlu: An Essay. Expedition 31/23: 107127.
Dyson, R.H., Jr., and Muscarella, O.W.: Constructing the Chronology and Historical
Implications of Hasanlu IV. Iran 27: 127.
Marcus, M.I.: Emblems of Authority: The Seals and Sealings from Hasanlu IVB.
Expedition 31/23: 5363.
Muscarella, O.W.: Warfare at Hasanlu in the Late 9th Century bc Expedition 31/23:
2436.
Pigott, V.C.: The Emergence of Iron Use at Hasanlu. Expedition 31/23: 6779.
Reese, D.S.: Treasures from the Sea: Shells and Shell Ornaments from Hasanlu.
Expedition 31/23: 8086.
Winter, I.J.: The Hasanlu Gold Bowl: Thirty Years Later. Expedition 31/23: 87
106.
1990
Marcus, M.I.:
1990a: Centre, Province and Periphery: A New Paradigm from Iron-Age Iran. Art
History 13/2: 129150.
1990b: Glyptic Style and Seal Function: The Hasanlu Connection. Aegeum 5: 175
193.
1991
Marcus, M.I.: The Mosaic Glass Vessels from Hasanlu, Iran: A Study in Large-Scale
Stylistic Trait Distribution. Art Bulletin 73: 536560.
1993
Marcus, M.I.: Incorporating the Body: Adornment, Gender, and Social Identity in
Ancient Iran. Cambridge Archaeological Journal 3: 157178.
1994
Marcus, M.I.:
1994a: In His Lips He Held a Spell. Source 13/4: 914.
1994b: Dressed to Kill: Women and Pins in Early Iran. Oxford Art Journal 17/2: 315.
Muscarella, O.W.: North-western Iran: Bronze Age to Iron Age. Pp. 139155 in Anato-
lian Iron Ages 3: The Proceedings of the Third Anatolian Iron Ages Colloquium Held
at Van, 612 August 1990, eds. A. ilingiroglu and D.H. French, London: British
Institute of Archaeology at Ankara.
Young, T.C., Jr.: Architectural Developments in Iron Age Western Iran. Bulletin of the
Canadian Society for Mesopotamian Studies 27: 2532.
1995
Muscarella, 0. W.: Art and Archaeology of Western Iran in Prehistory. Pp. 981999 in
Civilizations of the Ancient Near East, Vol. 2, ed. J.M. Sasson. New York: Scribners.
1996
Marcus, M.I.:
1996a: Emblems of Identity and Prestige: The Seals and Seatings from Hasanlu, Iran:
Commentary and Catalog. University Museum Monograph 84; Hasanlu
Special Studies 3. Philadelphia: University Museum, University of Pennsyl-
vania.
1996b: Sex and the Politics of Female Adornment in Pre-Achaemenid Iran (1000
800B.C.E.), Pp. 4154 in Sexuality in Ancient Art: Near East, Egypt, Greece,
and Italy, ed. N.B. Kampen. Cambridge: Cambridge University.
Muscarella, O.W.: Hasanlu. Pp. 209211 in The Dictionary of Art, Vol. 14, ed. J. Turner.
New York: Grove.
346 chapter nine
1997
Dyson, R.H., Jr.: Hasanlu. Pp. 478481 in The Oxford Encyclopedia of Archaeology in
the Near East, Vol. 2, ed. E.M. Meyers. New York: Oxford University.
1999
Dyson, R.H., Jr.:
1999a: Triangle-Festoon Ware Reconsidered. Iranica Antiqua 34: 115144.
1999b: The Achaemenid Painted Pottery of Hasanlu IIIA. Anatolian Studies 49: 101
110.
2001
de Schauensee, M.: A Note on Three Glass Plaques from Hasanlu. Iraq 63: 99106.
2002
Young, T.C., Jr.: Syria and Iran: Further Thoughts on the Architecture of Hasanlu.
Pp. 386398 in Of Pots and Pans: Papers on the Archaeology and History of Meso-
potamia and Syria Presented to David Oates in Honour of his 75th Birthday, eds. L.
al-Gailani Werr, J. Curtis, H. Martin, A. McMahon, J. Oates, and J. Reade. London:
NABU.
2003
Dyson, R.H., Jr.: Hasanlu Teppe. Pp. 4146 in Encyclopaedia Iranica, Vol. 12, ed.
E. Yarshater, Boston: Routledge and Kegan Paul.
Dyson, R.H., Jr., and Voigt, M.M.: A Temple at Hasanlu. Pp. 219236 in Yek bud,
yeki nabud: Essays on the Archaeology of Iran in Honor on William H. Sumner,
eds. N.F. Miller and K. Abdi. Monograph 48. Los Angeles: Cotsen Institute of
Archaeology, University of California.
2004
Danti, M.D.: The Ilkhanid Heartland: Hasanlu Tepe (Iran) Period I. University Muse-
um Monograph 120; Hasanlu Excavation Reports 2. Philadelphia: University of
Pennsylvania Museum of Archaeology and Anthropology.
Danti, M.D.; Voigt, M.M.; and Dyson, R.H., Jr.: The Search for the Late Chalcol-
ithic/Early Bronze Age Transition in the Ushnu-Solduz Valley, Iran. Pp. 584
616 in A View from the Highlands: Archaeological Studies in Honour of Charles
Burney, ed. A. Sagona. Ancient Near Eastern Studies, Supplement 12. Leuven:
Peeters.
Muscarella, O.W.: The Hasanlu Lion Pins Again. Pp. 693710 in A View From the
Highlands: Archaeological Studies in Honour of Charles Burney, ed. A. Sagona.
Ancient Near Eastern Studies, Supplement 12. Leuven: Peeters.
*Supplementary References
Barnett, R.D.
1982 Urartu. Pp. 314371 in The Cambridge Ancient History, Vol. 3, Pt. 1: The
Prehistory of the Balkans; and the Middle East and the Aegean World, Tenth
to Eighth Centuries bc, eds. J. Boardman, I.E.S. Edwards, N.G.L. Hammond,
and E. Sollberger. 2nd ed. Cambridge: Cambridge University.
Calmeyer, P.
1975 Hasanlu, Reallexikon der Assyriologie und Vorderasiatischen Archologie 4:
128131.
Dandamaev, M.A., and Lukonin, V.G.

1989 The Culture and Social Institutions of Ancient Iran. Cambridge: Cambridge
University.
Derin Z., and Muscarella, O.W.

2001 Iron and Bronze Arrows. Pp. 189217 in Ayanis I: Ten Years Excavations at
Rusahinili Eiduru-kai, 19891998, eds. A. ilingiroglu and M. Salvini. Docu-
menta Asiana 6. Rome: Istituto per gli Studi Miceni ed Egeo-Anatolici.
Dittmann, R.
1990 Eisenzeit I und II in West- und Nordwest-Iran zeitgleich zur Karum-Zeit
Anatoliens? Archologische Mitteilungen aus Iran 23: 105138.
Dyson, R.H., Jr.

1969 Preliminary Report on Work Carried out during 1968 by the Hasanlu Project
in Azerbaijan. Bastan Chenassi va Honor-e Iran 2: 1519.
1973 The Archaeological Evidence of the Second Millennium bc on the Persian
Plateau. Pp. 686715 in The Cambridge Ancient History, Vol. 2, Pt. 1: History
of the Middle East and the Aegean Region c. 18001380bc, eds. I.E.S. Edwards,
C.J. Gadd, N.G.L. Hammond, and E. Sollberger. 3rd ed. Cambridge: Cam-
bridge University.
Dyson, R.H., Jr.; Muscarella, O.W.; and Voigt, M.M.

1969 Hasanlu Project 1968: Hajji Firuz, Dinkha Tepe, Se Girdan, Qalatgah. Iran 7:
179181.
Fehr, B.
19711972 Zur Geschichte des Apollonheiligtums von Didyma. Marburger Win-
ckelmann-Programm 1971/72: 1459.
Ghirshman, R.
1939 Fouilles de Sialk prs de Kashan 1933, 1934, 1937, Vol. 2. Srie archologique 5.
Paris: Geuthner.
348 chapter nine
Ivantchik, A.I.
2001 Kimmerier und Skythen. Steppenvlker Eurasiens 2. Moscow: Eurasien-
Abteiling. des Deutschen Archaologischen Instituts.
Johnson, E.C.
1976 Urartian Influence upon Iranian Architecture in the Early First Millennium
bc Marsyas 17: 2138.
Kleiss, W.
1973 Planaufnahmen Urartischer Burgen in Iranisch-Azerbaidjan im Jahre 1972.
Archaeologische Mitteilungen aus Iran 6: 8189.
Kroll, S.
1975 Ein Schssel der Triangle Ware aus Azarbaidschan. Archaeologische Mit-
teilungen aus Iran 8: 7174.
1976 Keramik urartischer Festungen in Iran. Archologische Mitteilungen aus
Iran, Ergnzungsband 2. Berlin: Reimer.
1977 [Urartische Pltze in Iran]. B: die Oberflchenfunde des Uranu-Surveys
1976. Archaeologische Mitteilungen aus Iran 10: 83118.
1992 Ein triple road system oder Stallbauten in Hasanlu. Archologische Mit-
Levine, L.
1987 The Iron Age. Pp. 229250 in The Archaeology of Western Iran: Settlement
and Society from Prehistory to the Islamic Conquest, ed. F. Hole. Washington,
DC: Smithsonian Institution.
Levine, L.D., and Young, T.C., Jr., eds.

1977 Mountains and Lowlands: Essays in the Archaeology of Greater Mesopotamia.
Bibliotheca Mesopotamica 7. Malibu: Undena.
Marcus, M.I.
1995 Art and Ideology in Ancient Western Asia. Pp. 2487505 in Civilizations of
the Ancient Near East, Vol. 4, ed. J.M. Sasson. New York: Scribners.
Medvedskaya, I.
1982 Iran: Iron Age I. Trans. S. Pavlovich, from Russian. BAR International Series
126. Oxford: B.A.R.
Mousavi, A.
2005 Comments on the Early Iron Age in Iran. Iranica Antiqua 40: 8799.
Muscarella, O.W.
1968 Excavations at Dinkha Tepe, 1966. Metropolitan Museum of Art Bulletin 27/3:
187196.
1969 The Tumuli at S Girdan: A Preliminary Report. Metropolitan Museum
Journal 2: 525.
1971a Qalatgah: An Urartian Site in Northwestern Iran. Expedition 13/34: 44

49.
1971b The Tumuli at S Girdan: Second Report. Metropolitan Museum Journal 4:
528.
1973 Excavations at Agrab Tepe. Metropolitan Museum Journal 8: 4776.
1974 The Iron Age at Dinkha Tepe, Iran. Metropolitan Museum Journal 9: 3590.
1988 Bronze and Iron: Ancient Near Eastern Artifacts in The Metropolitan Museum
of Art. New York: Metropolitan Museum of Art.
2003 The Chronology and Culture of S Girdan: Phase III. Ancient Civilizations
from Scythia to Siberia 9: 117131.
Porada, E.
1965 The Art of Ancient Iran: Pre-Islamic Cultures. New York: Crown.
Reade, J.
1979 Hasanlu, Gilzanu, and Related Considerations. Archaeologische Mitteilun-
gen aus Iran 12: 175181.
Renfrew, C.
1980 The Great Tradition versus the Great Divide: Archaeology as Anthropology?
American Journal of Archaeology 84: 287298.
Salvini, M.
1995 Geschichte und Kultur der Urarter. Darmstadt: Wissenschaftliche Buchge-
sellschaft.
Stllner, T.; Slotta, A.; and Vatandoust, A.

2004 Persiens antike Pracht: Bergbau, Handwerk, Archologie: Katalog der Ausstel-
lung des Deutschen Bergbau-Museums Bochum vom 28. November 2004 bis
29. Mai 2005. Bochum: Deutsches Bergbau-Museum.
van Loon, M.
1975 The Inscription of Ishpuini and Meinua at Qalatgah, Iran. Journal of Near
Eastern Studies 34: 201207.
Young, T.C., Jr.

1967 The Iranian Migration into the Zagros. Iran 5: 1134.
Zimansky, P.
1995 Urartian Material Culture as State Assemblage: An Anomaly in the Archae-
ology of Empire. Bulletin of the American Schools of Oriental Research 299/
300: 103115.
chapter ten
THE IRANIAN IRON III CHRONOLOGY

AT MUWEILAH IN THE EMIRATE OF SHARJAH*
Abstract: The site of Muweilah in Sharjah, United Arab Emirates, has been pub-
lished by its excavator Peter Magee over a number of years as having flourished
during the Iron II period of the UAE/Oman, Arabian chronological system, ca. 1100/
1000600bc. He has further asserted that within this long period, Muweilahs exis-
tence can be dated to the time of the north-western Iranian Iron II period, which
terminated ca. 800bc, dating his site specifically to ca. 920770 bc. Evidence used
to affirm the Iranian Iron II chronology includes Iranian and local architecture
and pottery parallels, and C 14 data. I rejected the viability and relevance of these
parallels in print in 2003, to which Magee responded, reaffirming his 10th-early 8th-
century bc chronology. Here I respond to the excavators ongoing defence, and argue
for a considerably later north-western Iranian Iron III date for Muweilah.
The Background
Reacting to two of Peter Magees articles (of 1997 and 2001) about his site
Muweilah in Sharjah, United Arab Emirates, I challenged the chronology he
presented, claiming it was too high, and disagreed with his cultural/histori-
cal conclusions relating to the nature and date of the pottery and architec-
ture recovered.1 Magee responded to my challenge,2 vigorously defending
(appropriately) his position. Here I offer my response to his 2005 defence:
I have not changed my mind regarding his dating of Muweilah and present
* This article originally appeared as The Iranian Iron III Chronology at Muweilah in the
Emirate of Sharjah, Ancient West and East 7 (2008): 189202.

I want to thank Ernie Haerinck and Dan Potts for making suggestions, suggesting
bibliography, and sharing their views with me. And Peter Magee for graciously allowing me
to publish photographs from his publications.
1 Muscarella 2003, 249250, n. 102. My critique was presented briefly in a footnote in an
article a new C 14 date recently proclaimed by the excavators of Gordion in Anatolia, arguing
that it was too high. The footnote was presented to give another example of a C 14 date that I
believed to be incorrect. I first encountered Muweilah at a Bryn Mawr lecture by Peter Magee
in October 2002, where I first expressed (verbally) my disagreements about the chronology
he assigned to Muweilahs and architecture.
2 Magee 2005a.
352 chapter ten
my arguments for rejection here. Beginning with his first reports Magee
has continuously reported that Muweilah came into existence during the
UAE/Oman, Arabian Iron II period, which he dates from ca. 1100/1000 to
600bc.3 The reader must understand from the beginning of the discussion
that this chronological period, Iron II, is alleged to have lasted for 500 years;
and that in Iran this same 500-year time-span encompasses two separate
and distinct cultural and chronological periods, Iron II and Iron IIIcrucial
issues not articulated by Magee. Within that broad Arabian Iron II time
frame he specifically situates Muweilahs construction and existence con-
temporary with the chronology of the Iranian Iron II period, with Hasanlu
Period IV, which terminated ca. 800 bc (i.e. 200 years earlier than the appar-
ent termination of the Iron II period in Arabian terminology).4
The Ceramics
One significant component of the collective evidence he presents is that

Muweilah has vessels with bridged spouts (bridged describes the unit that
connects the spout to the rim), which to him are forms typical of the Iranian
Iron II period (Fig. 1).5 However, vessels with a bridged horizontal spout are a
classic characteristic form of the Iranian Iron II period (Fig. 3). And although
many of the spouts of the published Muweilah examples are broken-away,
some are intact and reveal, not a horizontal but an upright, vertical spout
a characteristic not of the Iranian Iron II period, but of Iron III. He states
that some Muweilah examples have a short bridge, others have no bridge.
But Muweilah has no typical Iranian Iron II-form horizontal spouts.
Magee also presents references and drawings of vessels from Rumeilah, a
nearby site, as relevant comparanda for his asserted UAE and Iranian Iron II
chronology there also.6 None has a bridged horizontal spout, and the exam-
ples presented are painted (as at Muweilah7Fig. 2), which decoration is a
3 Magee 1996a, 208; 1996b, 246, 249; 1997, 96; 1999, 44; 2002, 161; 2004, 32; 2005a, 161; 96.
4 I am not a scholar of Arabian archaeology and I found it confusing that the very same
terminology used in Iranian archaeology, Iron I, II, III, is employed in Arabia. For a discussion
of the Iranian/Hasanlu Iron Age terminology and problems, see Muscarella 2006.
5 Magee 1996a, 203, 205206, figs. 1617; 1999, 45, figs. 56; 2001, 121, 123, fig. 12; 2002, 164
165, fig. 2; 2005a, 165, fig. 1, right (compare the Hasanlu vessel at the right); 2005b, 99, 112, figs. 5
(Fig. 1 in this paper), 20; Magee et al. 2002, 141, fig. 13.
6 Magee 1996a, 208; 1996b, 246248, fig. 7.B, C, E, G; 1997, 9395, fig. 2; 2005a, 165, fig. 1,
centre; see also Boucharlat and Lombard 2001, 218, fig. 11.
7 Magee 1999, 45, fig. 5; 2005b, 100, fig. 6.
the iranian iron iii chronology at muweilah 353
Figure 1. Spouted vessels from Muweilah.

354 chapter ten
Figure 2. Assorted vessels from Muweilah.

Figure 3. Bridged horizontal spouted vessel from Hasanlu IV.

The Metropolitan Museum of Art, Rogers Fund, 1960
(65.163.72) Image The Metropolitan Museum of Art.
manifest post-Hasanlu IV/Iranian Iron II characteristic (below); and formal

parallels for the spout form of E, joined to and level with the vessel rim, are
from Luristan in the Luristan Iron III period.8
Sialk B (a cemetery) has painted bridged horizontal spout vessels, which
were cited by Magee as Iranian Iron II chronological parallels for Rumeilah,
and later also cited as evidence supporting his early Muweilah chronology.9
He quotes Dyson, who, discussing the spout-painted juxtaposition there,
claimed: The main occupation of Sialk B would then belong to the eighth
century overlapping the end of Hasanlu IVB and the beginning of IIIB.10
The problem here is that Dyson made a significant error: painted pottery in
fact does not appear in Hasanlu IV or in the following period Hasanlu III B,
the Urartian period, but after the latters destruction, in Period III A, which
came into existence not earlier than the late 7th or early 6th century bc.11
8 Overlaet 2005, pls. 8.2, 9, 11.34. Further, the drawing of B (in n. 5 above) seems to be a
restoration.
9 Magee 1997, 94, 96, fig. 2; 1999,45; 2005a, 162; 2005b, 9394.
10 Dyson 1965, 208. On this also, see Young 1965, 7680, figs. 1314.
11 This reality has been known for decades: viz. Haerinck 1978, 85. For details and bibliog-
raphy, see Muscarella 2006, 15, 1720.

356 chapter ten
I too would agree with Magee that it is too early to make definitive
statements concerning the date of Iron Age Sialk,12 which is a complicated
subject. But I would add, that based on the Sialk pottery, one cannot date
Rumeilah (or, indeed, Muweilah) within the Iranian Iron II chronology as
established in north-western Iran. Magee himself admits (but then ignores
it) that the Arabian decorative patterns are unlike the examples found in
Iran during this period, i.e. in the context here, Iron II.13
Magee continuously uses the terms bridged or bridged-spouted to
describe the Muweilah spout form, but inexplicably does not inform us
whether the bridged spouts are positioned vertically or horizontally. The
position of the spout determined how the liquid was poured out into
another container, and although Magee mentions this process, he does not
inform us how far the Muweilah people had to tip their spouts.14 Spouted
vessel forms from the Iranian Iron III period exist, viz. at Hasanlu, Ziwiye
and Yanik Tepe.15
In response to my 2003 brief comments on this significant matter, Magee
replies16 it is difficult to know how Muscarella can be certain that they
are horizontally or vertically spouted. My answer is that lacking textual
information, I looked at his published photographs, which do not portray
horizontal bridged spouts; and how and why does Magee know that they
are horizontally spouted: which must be the case to qualify for his Iron II
attribution? But nowhere has he addressed this, recognised its significance
in his discussions of parallels and chronology. He subtly adds two ambigu-
ous modifications to his previously published claims.17 One is that not all
the Muweilah examples are comparable to the more elongated horizontal
bridge-spouted examples from north-western Iran. If by more elongated
he is now stating that the (some?) spouts at Muweilah are bridged horizon-
tally, but are short, why not say it straightforward? Note that in this 2005
article the horizontal word for the Muweilah vessels is mentioned for the
first time, again ignoring its chronological importance.18 He also claims that
12 Magee 2005a, 163.

13 Magee 2005b, 99.
14 Magee 2005b, 108109. For the spout pouring position by the Iranian Iron II population,
see Stein 1940, pl. XXX, 8.

15 Hasanlu: Young 1965, 56, fig. 2.1 (the spout is incorrectly restored; it is vertical, as
recognised by Haerinck 1987, 87); Ziwiye: Young 1965 60, fig. 4.5; Yanik Tepe: Haerinck 1978,
84, 87, fig. 7.
16 Magee 2005a, 163164.
17 Magee 2005a, 164.
18 Magee 2005a, 164.
he is not able to see how any of the Muweilah examples are significantly
[his emphasis] different from Iranian Iron Age II bridge-spouted vessels
in general [my emphasis]. In another venue published the same year19 he
also stated, obliquely and casually, that there are many differences between
the Arabian and west and northwest Iranian examples. The east Arabian
examples contain a more open spout that differs from the very elongated
horizontal [sic] spout found on some [sic] northwest and central Iranian
examples (no profile drawings are published). The vertical spout position
remains unmentioned here, concealed from us, although the vertical word
is used when informing us that vertical spouts of a different form exist at
Sialk. Here also he generalises about a long life for bridge-spouted vessels
down to the 8th century bc.20
I suggest that Magees 2005 defence contra Muscarella 2003 is an attempt
to modify all his previously published, and strongly stated, pottery parallel
claims. Here is the core of the issue under review: I see only bridged vertical
spouted vessels among those published from Muweilah (Figs. 12), a form in
Iran that is stratigraphically distinguished as occurring later than the Iron II
horizontal spout examples. In 1996 Magee asserted an indefinite generalisa-
tion, Bridge-vessels are leitfossils for the Iron II period in western Iran,21 but
omitted the qualifying horizontal, a term essential in any characterisation
of Iranian Iron II leitfossil spouted vessels. For excellent published Hasanlu
Period IV examples see Stein; for the record, there are examples of bridged
horizontal spout vessels in post-Iron II Iran.22 Magee23 also cites examples
of bridged spouted vessels at Godin and Baba Jan, and which (again) is
meaningless, and in the context a misleading use of bridged (why omit
the crucial position of the spout?). In fact, all the vessels from these sites
are dated to the Iranian Iron III periodand all the cited vessels have verti-
cal spouts; also, Baba Jan has much painted pottery:24 which as such should
19 Magee 2005b, 99.

20 Magee 2005b, 94.
21 Magee 1996b, 248.
22 Stein 1940, pls. XXIV, XXX, no. 8; compare the Iron III examples there, pl. XVII, and the
drawing in pl. XXVIII, 17. An excavated example from Ziwiye has hanging triangles and incised
chequerboard decoration (Dyson 1965, 206, fig. 11): it has a parallel at Muweilah (Magee
2005b, 113, n. 28); Nush-I Jan (Stronach 1969, 18, fig. 7); War Kabud in Lurisran (Overlaet 2005,
pl. 11.1), which may possibly be dated in Luristan chronology to Iron IIB, 8th century bc or
later.
23 Magee 2005b, 94.
24 Magee (2005a, 94, n. 9) cites Goff 1985 as a Baba Jan reference. This is an error, repeated
358 chapter ten
have been employed to date Muweilah (and Rumeilah, as below) to the

Iranian Iron III period.
Magee claims25 it was not [his emphasis] bridge-spouted vessels that
were the sole basis for his dating of Muweilah, it was the C 14 evidence (see
below). Indeed he unquestionably (and vitally; see below) depends on the
carbon dates. But, notwithstanding his disclaimer, no reader of Magees pub-
lications can be unaware that his Iranian Iron II pottery-parallel assertions
are a major component of his arguments concerning the specific cultural,
geographical source and chronology at his site.26 Note also his statement
that absolute chronology of the Iron II period is available from both C 14
data and [my emphasis] foreign parallels for Iron II pottery.27 Also claimed
here is that: The most chronologically diagnostic evidence [my emphasis] for
foreign inspiration in the Iron II period is found in painted and unpainted
bridge-spouted vessels.28 True. But there are no painted vessels in Iron II
Hasanlu, or Dinkha Tepe, or at any other manifest Iron II sites in north-
western Iran (above).
Also unrecognised by Magee is that there is more ceramic evidence avail-
able at Muweilah to support a post-Iron II occupation there, and by cen-
turies. First is a distinct vessel form called a Trichterrandschale, charac-
terised by a relatively long flaring rim joining a bulging body (Fig. 2, centre
and lower).29 There are no known Iranian Iron II paralells; they occur in
Iron III contexts, and later, in the Achaemenid period: see the long list of
post Iranian Iron II sites where they occur given by Kroll.30
Then there are fragments of two terracotta vessels identified by him as
incense burners that are decorated in typical Iron Age II fashion: swirling
lines around the holes.31 The example illustrated has two isolated holes,
and does not appear to be a censor; furthermore, he gives no parallels to
support the typical Iron II attribution. Also recovered is a domed, fully
in the bibliography: the journal reference should be Goff 1978, and the issue is Iran 16, pp. 29
66 (for a painted vessel with vertical spout, see pl. 1Ia). Nn. 6 and 7 also have wrong references.
The Manor house at Baba Jan does nor seem to have a columned hall; it has three columns,
probably for a colonnade. For the post-9th centuty dating of Baba Jan, see Muscarella 1988,
140, n. 1, 209, n. 4 (contra Boucharlat and Lombard 2001, 222, n. 9).
25 Magee 2005a, 163.
26 See Magee 1996a, 208; 1997, 9395 (Rumeilah); 2001, 123; 2002, 164.
27 Magee 1996b, 247.
28 Magee 1996b, 247.
29 Magee 2005b, fig. 7, centre.
30 Kroll 1976, 115; see also Young 1965, 58, fig. 3.6, 11, fig. 4.6.
31 Magee 2004,2728, fig. 4 for one example.
Figure 4. Ceramic censor cover from Muweilah.
perforated manifest ceramic incense burner cover crowned by a figure of

a bull (Fig. 4).32 Magee informs us that this censor fits within the reper-
toire of Iron II ceramics, and says parallels occur at three Arabian sites;
Magee is here employing Iron II only in its UAE/Arabian Iron II terminol-
ogy (ca. 1100/1000600 bc, see above). But the censor cover most certainly
cannot be used as evidence to support a 9th-century bc Iranian Iron II date
for Muweilah: because there are no Iranian Iron II parallels, there are only
post Iranian Iron II parallels. B. Goldman has conveniently brought most
known censors together (Arabian examples are not mentioned).33 Censors
on stands begin in the 2nd millennium bc, but rounded, removable domed
covers are rare, only that depicted on a Hittite scene (GoldmanCC), and
32 Magee 2001, 123124, fig. 14; 2005b, 112113, fig. 21.

33 Goldman 1991.
360 chapter ten
it is not perforated. The next rounded cover example known is from the 7th
century bc. Ashurbanipal garden relief (GoldmanLL), also unperforated.
Only those from the Achaemenid period have all the characteristics of the
Muweilah censor cover: form, multi-perforations and, sometimes, figured
handles (GoldmanA, B, F, G, I).34 Some are domed, others pyramid shaped
and stepped, for which also see the Achaemenid examples from Usak.35 One
can argue that the Achaemenid-form censor derives from a late phase at
Muweilah, the time prior to its destruction, but then indeed this could also
obtain for the vertically spouted vessels as well: which were recovered in
the destruction level there.36 The archaeological evidence indicates that the
Muweilah censor cover cannot be employed to date the sites destruction to
a fire sometime after 770 bc but before c. 600 bc.37 Its presence contradicts
this asserted chronology: and brings into the discussion an Achaemenid
period at Muweilah (whether it can be argued that such censors existed
earlier in Arabia, I leave to the specialists); Magee plays down involvement
with Fars.38
The Architecture
Architecture is another major component of Magees Iranian Iron II chrono-

logical evaluation. The vital argument here is that Muweilahs columned
hall plan is crucial evidence for comprehending a date for the site. Because
Magee consistently perceives this plan as distinctly related to Burnt Build-
ing II (BB II) at Hasanlu, Period IV (Fig. 5), it thus signifies contemporary
existence, a chronological parallel and dependence; and inasmuch as BB II
is earlier than those at Muweilah one must assume [my emphasis], there-
fore, that influence was exerted from Iran to southeastern Arabia at this
time39 (10th9th centuries bc; see below). The architectural parallels further
signify to him that there must have been direct and complex social contacts
between the two widely distant sites.40 Such assumptions have no empirical
archaeological or historical support in Arabia or north-western Iran.
34 Goldman 1991, pl. XVII.

35 zgen and ztrk 1996, 114119, nos. 7173.
36 Magee 2005b, 98.
37 Magee 2001, 115; see also below.
38 Magee 2005b, 107, 112.
39 Magee 2002, 162163; see Magee 2001, 117, fig. 2; 2002, 165; 2005a, 167, fig. 4; Magee et al.
2002, 138, fig. 6.

40 Magee 2001, 128; 2005a, 164.
Figure 5. Columned Hall plans of Hasanlu,

Muweilah, Nush-i Jan and Godin.
362 chapter ten
Muweilahs columned hall is best described as an apadana (whether with

stone or wood columns is irrelevant), and has closer formal parallels at
geographically closer sites in western Iran. The hall has twenty columns
bases arranged in a five by four pattern with one row against the wall
that fill the interior space.41 Hasanlu BB II has eight central columns, four
each in two rows, with space all around them, and a series of smaller
columns situated against the walls; other buildings at Hasanlu have eight
or four central columns.42 The two sites have different column numbers and
internal dispositions.
Although Magee does mention the columned halls at the late 8th7th
century bc sites at Nush-i Jan and Godin Tepe (also Persepolis and Pasar-
gadaei),43 which I had cited as appropriate parallels (Fig. 5), closer in plan to
Muweilah, and that the sites are geographically closer than Hasanlu, they are
dropped from further investigation (until 2005): because, I suggest, Hasanlu
BB II had been accepted as the Muweilah parallel. In 2005 my judgment
that Hasanlu BB II is not a suitable parallel for Muweilah, which parallel is
more meaningfully recognised at the laterdated sites, was rejected.44 I read-
ily reaffirm it. Nush-i Jan has a twelve-columned hall, three rows of four
each; Godin has a central hall with thirty columns, five rows of six each, and
two flanking narrower halls with a smaller number of columns (10, 16; partly
reconstructed). Magee leaves it to the reader to decide, but is immediately
compelled to add that the Hasanlu-Muweilah parallels seem to me stronger
than those from Nush-i Jan and Godin. I reject the urgency of this claim.
To me the Muweilah building is paralleled by the apadana plans, which
are kindred in their column number, placements and concept. BB IIs plan
cannot be architecturally and formally privileged over those from Iranian
Iron III Nush-i Jan and Godin.45 Boucharlat and Lombard46 also discuss these
two sites, accepting their filiation with Rumeilah (they do not mention
Muweilah!) as not inconceivable, and accept them all being contemporary.
Note also that a possible columned hall has been partly excavated in Iran,
at Ziwiye: a room with sixteen columns in two rows was uncovered, but this
may be a colonnade.
41 Magee 2002, 162.

42 Dyson 1965, pl. XXXIV; for a full plan of the Period IV columned hall structures, see
Muscarella 2006, 9, fig. 3.
43 Magee 2002, 163.
44 Magee 2005a, 164.
45 The juxtaposition of the building plans under review is conveniently displayed in
Magee 2005a, fig. 4here Fig. 5.

46 Boucharlat and Lombard 2001, 221222.
Although a columned hall was excavated at Rumeilah,47 it gets but a mere

mention by Magee.48 Rumeilahs hall has three columns in three rows; an
underlying structure seems also to have had columns, although it is possi-
ble here we have a partially covered area.49 Boucharlat and Lombard date
Rumeilah within the Arabian Iron II period, but nota bene, they place its
chronological phase at the latest to the late 8th7th century bc.50 This is rea-
sonable, indeed, and must form the beginning discussion of the chronology
of Muweilah.
A relevant academic issue is worth raising: if indeed Muweilahs apadana
was constructed in the Iranian Iron II period, then, in this scheme, may
not one ponder if the Median examples in Iran could have been derived
from South Arabia, and not from an indigenous northern Iranian back-
ground? For, if Muweilah had contact with far-away Hasanlu in the 10th
9th centuries bc, why not consider a continuous contact with the main-
land, with the closer Median area, whose buildings were built after those at
Muweilah? Magee doesnt confront this view, but seems to obliquely reject
it.51 While casually accepting Magees chronology (Muweilah apparently
[sic] dates from before 800bc), Curtis and Razmijou52 correctly reject Ara-
bia as the source for the Median and Achaemenian apadanas. But they also
believe that the architecture in both areas experienced inspiration from the
same tradition, which, independently obtained, they argue, derived from
Hasanlu. Thus they support Magees chronological and cultural conclu-
sionsnorth-western Iranian influence and contact with Arabia in the Ira-
nian Iron II periodwhich my present paper again rejects.
Carbon-14 Data
From the earliest reports Magee cited C 14 data as evidence to situate

Muweilah chronologically in the Iranian Iron II Period: in 1999 the build-
ings were in existence by the ninth century bc destroyed sometime
after 770 bc;53 in 2001 the buildings came into existence sometime after
47 Boucharlat and Lombard 2001, 205218, figs. 3, 7.

48 Magee 2005b, 109.
49 Boucharlat and Lombard 2001, 215, fig. 6.
50 Boucharlat and Lombard 2001, 221, n. 9.
51 Magee 2002, 163.
52 Curtis and Razmijou 2005, 50.
53 Magee 1999, 4647.
364 chapter ten
ca. 920bc destroyed by fire 770bc but before c. 600 bc;54 in 2002 the initial
construction date is lowered to ca. 900 bc;55 in 2003 it is after 920 bc and with
an upper limit of 800 bc for the sites destruction (wood and date seeds).56
In 2004 the construction date for the site is now the end of the ninth cen-
tury bc, based on dates and date seeds samples.57 The following year Magee
emphasised the C 14 evidence from dates and burnt beams for the chronol-
ogy, repeating that the main buildings were constructed after ca. 920bc
and destroyed sometime after ca. 800 bc but before ca. 600bc.58 And bridge
spouted vessels were in use, therefore, sometime after 920bc, perhaps even
from c. 1000 bc onwards, and exhibit some form of influence from Iran [my
emphasis].59 In the final analysis, C 14 evidence is the core of the issue at
hand. Surely it was the C 14 dates that firmed his chronological claims, his
associated dating of the pottery and architecture. All three are presented as
if they naturally form a closed, self-created triangle of mutually supporting
evidence.
The C 14 dates given in Magee 2004, along with data given in table 1 there,
do not manifestly prove the Iranian Iron II chronology argued. Here I stick
my neck out, for I see a problem with the C 14 analysis; and I must leave a
close analysis to C 14 statistical experts to debate (as they will also with my
2003 article that disagrees with another C 14 determination).
One more issue, an important conclusion of Magees, should be ad-
dressed. Magee believes that Arabian letters inscribed on a vessel fragment
excavated at Muweilah manifest that the South Arabian script existed in
south-eastern Arabia, and thus, based on his chronology, indicates that seri-
ous cultural interaction with Yemen began 300 to 400 years earlier than
previously thought.60 If I am correct in my disagreements with Magees
Muweilahs chronology, this cultural assertion is erroneous, but warrants
independent discussion by the experts.
54 Magee 2001, 115.

55 Magee et al. 2002, 153154.
56 Magee 2003, 23, 7.
57 Magee 2004, 32.
58 Magee 2005a, 163; see also 2005b, 9798.
59 Magee 2005b, 9899.
60 Magee 1999, 4345, fig. 3, 4749.
Conclusions
Based then on the pottery, the censor and formal architectural parallels, and
a challenge of the C 14 date, I argue that Muweilah came into existence in the
post-Iranian Iron II period, probably in the 7th century bc (or later?), what-
ever UAE/Arabian terminology is employed (late Iron II?). When Magee
states that there is an absence of iron III material culture at Muweilah,61 he
can only be referring to Arabian terminology. And Muscarella never commit-
ted the fundamental error of only equating this pottery and thus this site.62
(One may aptly use the same phrase to describe Magees equating Hasanlu
with Muweilah.) What I said and repeat, is that no ceramic or architectural
evidence present at Muweilah, or Rumeilah, supports an Iranian Iron II date
there. The issue is not, and never was, whether there was Iranian influence
and trade with Arabia,63 but when it commenced.
Addendum
In the latest issue of BASOR (No. 347 [2007], 83105; it appeared at the
Metropolitan Museum of Art on 20th September 2007) Peter Magee has a
long article on Arabian sites (Beyond the Desert and the Sown ), including
Muweilah. Here in a long discussion on chronology he dates Muweilah
(along with other sites) generally within the Arabian Iron II period, with a
broad 500 year range date of 1100600 bc. But not once here does he mention
that in his many (and repetitious) earlier writings he has consistently dated
Muweilah to the 9th century bc, vigorously asserting alleged Hasanlu IV B,
9th-century bc ceramic and architectural parallels. These articles, eleven in
alland of which a total of five are omitted/ignored in his argument and
from the bibliographyare discussed and critiqued by me in the above,
in the present article. There is more. Nowhere in the BASOR article does
he mention or include in his bibliography my original challenge to his 9th-
century bc dating of Muweilah that appeared in AWE 2.2 (2003), n. 102. And
nowhere does he include a reference to his long article in AWE 4.1 (2005)
contra my 2003 note, wherein he challenged my late dating (hundreds of
years later than his dating), and vigorously defended the 9th-century date
for Muweilah that he has touted for a long time. Furthermore, he does not
61 Magee 2004, 33; also 2003, 1.

62 Magee 2005a, 161.
63 Magee 2005b, 9596.
366 chapter ten
mention that he had read with my permission (upon the request of AWEs
Editor-in-Chief) in August 2006 the then unpublished article submitted to
AWEthe present articlerebutting his 2005 AWE paper!
Bibliography
Boucharlat, R. and Lombard, P. 2001: Le Btiment G de Rumeilah (Oasis dal Ain).

Remarques sur les Salles Poteaux de l Age du fer en Pninsule dOman. IranAnt
36, 213238.
Curtis, J. and Razmijou, S. 2005: The Palace. In Curtis, J. and Tallis, N. (eds.), Forgot-
ten Empire The World of Ancient Persia (London), 5055.
Dyson, R.H. jr 1965: Problems of Prehistoric Iran as seen from Hasanlu. JNES 24,
193217.
Goldman B. 1991: Persian Domed Turibula. StudIran 20, 179188.
Haerinck, H. 1978: Painted Pottery of the Ardebil Style in Azerbaidjan (Iran).
IranAnt XIII, 7591
Kroll, S. 1976: Keramik Urartischer Festungen in Iran (Berlin).
Magee, P. 1996a: Excavations at Muweilah. Preliminary Reports on the First Two
Seasons. Arabian Archaeology and Epigraphy 7, 195213.
1996b: The Chronology of the Southeast Arabian Iron Age. Arabian Archae-
ology and Epigraphy 7, 240252.
1997: The Iranian Iron Age and the Chronology of Settlement in Southeastern
Arabia. IranAnt32, 91108.
1999: Writing in the Iron Age: the earliest South Arabian inscription from
southeastern Arabia. Arabian Archeology and Epigraphy 10, 4350.
2001: Excavations at the Iron Age Settlement of Muweilah 19972000. Pro-
ceedings of the Seminar for Arabian Studies 31, 115130.
2002: The Indigenous Context of Foreign Exchange between South-eastern
Arabia and Iran in the Iron Age. The Journal of Oman Studies 12, 161168.
2003: New chronometric data defining the Iron Age period in southeastern
Arabia. Proceedings of the Seminar for Arabian Studies 33, 110.
2004: The impact of southeast Arabian intra-regional trade on settlement
location and organization during the Iron Age II period. Arab Archaeology and
Epigraphy 15, 2442.
2005a: Columned Halls, Bridge-Spouted Vessels, C 14 Dates and the Chronol-
ogy of the East Arabian Iron Age: A Response to Some Recent Comments by
O. Muscarella in Ancient West & East. In AWE 4.1, 160169.
2005b: The production, Distribution and Function or Iron Age Bridge-
Spouted Vessels in Iran and Arabia: Results from Recent Excavations and Geo-
chemical Analysis. Iran 43, 93115.
Magee, P., Thompson, E., Mackay, A., Kottaras, P. and Weeks, L. 2002: Further
evidence of desert settlement complexity: report on the 2001 excavations at the
Iron Age site of Muweilah, Emirate of Sharjah, United Arab Emirates. Arabian
Archaeology and Epigraphy 13, 133156.
Muscarella, O.W. 1988: Bronze and Iron (New York).
2003: The date of the Destruction of the Early Phrygian Period at Gordion.
AWE 2.2, 225252.
2006: The Excavation of Hasanlu: an Archaeological Evaluation. BASOR 342,
6994.
Overlaet, B. 2005: The Chronology of the Iron Age in the Pusht-i Kuh, Luristan.
IranAnt 40, 133.
zgen, T. and ztrk J. 1966: The Lydian Treasure (Istanbul).
Stein, A. 1940: Old Routes of Western Iran (London).
Stronach, D. 1969: Excavations at Tepe Nush-i Jan, 1967. Iran 7, 120.
Young, T.C. jr 1965: A Comparative Ceramic Chronology for Western Iran. 1500
500BC. Iran III, 5385.
chapter eleven
THE LOCATION OF ULHU AND UISE

IN SARGON IIS EIGHTH CAMPAIGN, 714 BC*
Abstract: In 714 bc, Sargon II, king of Assyria, conducted a major military campaign
across the Zagros Mountains into western Iran. It was the eighth campaign of his
reign, and the fourth into Iran. This time, Sargons main goal was to contain the
state of Urartu within its territory and to subdue its allies. The itinerary of the
Assyrian army and the events of the campaign were written in the form of a long
letter from the king to the god Assur, and it is preserved almost in its entirety. The
text is of great importance because it gives one of the most extensive itineraries of
an Assyrian campaign and, unlike other Assyrian reports, it mentions by name, and
often describes, a large number of cities and fortresses as well as many geographical
features.
For over 70 years scholars have attempted to reconstruct the route of Sargon from
Assyria to Iran and to identify by epigraphical and archaeological research and
survey the cities and features he mentions. Unfortunately, there has been little
agreement among the modern researchers with regard to the specific directions
taken by Sargon and equally so with attempts to link a site on the archaeological
map with one mentioned in the ancient text. The present article gives a summary
of the various solutions offered over the years and argues for the identification of
two archaeologically known sites with two of those mentioned by Sargon.
Introduction
The study of the historical geography of ancient Iran is laden with difficulties
and lack of verification. Although Mesopotamian cuneiform texts often
refer to specific districts, states, cities, mountains, and rivers in western
Iran, it is not possible by some easy methodology to equate with certainty
the toponyms with actual geographical loci. The main reason for this gap
between text and place is the absence from excavated sites in western Iran
of textual material that might furnish information about their names. In
the neo-Assyrian period of Mesopotamian history (10th7th centuries bc)
many texts exist that refer to and cite scores of features and places in western
* This article originally appeared as The Location of Ulhu and Uise in Sargon IIs Eighth
Campaign, 714bc Journal of Field Archaeology 13, no. 4 (1986): 465475.

370 chapter eleven
Iran, and many valiant attempts have been made by scholars to identify and
locate them. The major text of Sargon IIs eighth campaign gives an account
of the kings conquests and collection of booty in areas of western Iran
and contains quite specific citations and descriptions of both geographical
features and inhabited cities and fortresses. Attempts to link up or equate
the toponyms of the text with archaeological sites and geographical features
are based on the examination of local topography and site placement in
western Iran. But the various conclusions and identifications presented by
different scholars, using the same text and examining the same topography,
manifestly demonstrate that the problems of identification are not readily
resolved.
If one is able, therefore, to identify with some degree of certainty and
viability a single known Iranian site that is mentioned in Sargons text, it
becomes possible to renew the investigation with a fresh approach. Replac-
ing a modern mound-name with its ancient name would not only give some
indication of direction for Sargons journey from Assyria, but it might also, by
reference both to the sequence of movement supplied in the text and to local
archaeological features, allow further site identifications to be obtained. It
is the conclusion of this paper that, by utilizing epigraphical and geographi-
cal data and the results of recent archaeological excavation and survey, two
of the ancient fortress cities mentioned in Sargons text may with some cer-
tainty be identified with two known sites in nw Iran.
Earlier Interpretations
In the course of reviewing Paul Zimanskys book Ecology and Empire: The
Structure of the Urartian State (1985) I was obliged to consult in some detail
the text recording the eighth campaign of the Assyrian king Sargon II in
western Iran and in Urartian territory in 714 bc (Thureau-Dangin 1912; Meiss-
ner 1922; Luckenbill 1927:7379; Weidner 19371939). While Zimansky is not
primarily concerned with the labyrinthine problems associated with mod-
ern reconstructions of Sargons route, he engages these issues when they
touch upon the terminology employed in the Assyrian text to describe the
physical characteristics of Urartian settlements. Thus, when discussing the
great fortress (birtu rab) of Uaiais (line 299 of the Assyrian text), Ziman-
sky (1985: 42, 112, note 64) forcibly states that Sargons description of the site
fits well with Qalatgah. This equation, linking the important city of Uaiais
mentioned by Sargon with Qalatgah, an Urartian site in the Ushnu Valley
just sw of Lake Urmia in nw Iran (Azerbaijan), intrigued me, and the vigor
the location of ulhu and uise in sargon iis eighth campaign 371
of the claim compelled me to review the evidence that would allow such an
identification. This entailed a close reading of Sargons text and an exam-
ination of the archaeological data available from Qalatgah itself and from
the neighboring terrain. The results of this review, which led me to disagree
with Zimanskys identification, generated the writing of this paper.
The longest account of the eighth campaign of Sargon is written in cunei-
form on a large clay tablet (37.5 24.5cm) that derived from the Assyrian
capital city of Assur in Iraq. It was pilfered from that site and presumably
placed in the antiquities market; it eventually was acquired by the Louvre
zu Unrecht (Meissner 1922:113). The text was first published with commen-
tary by F. Thureau-Dangin in 1912, at a time when very little was known
about the archaeology and ancient cultural geography of Anatolia and Iran.
What is certain from the context of the text is the general direction of Sar-
gons route: that he crossed the Lower Zab River in Mesopotamia and passed
over Mt. Kullar into the Zagros Mountains and Iran, where he campaigned
before turning north to attack Urartian-controlled territory. After at least
one major battle was fought and a number of cities subdued, some near
a sea, Sargon returned home to Assyria. His army returned by one route,
while Sargon with 1000 picked troops went by another, and difficult, route
to attack Musasir, a holy city in ne Iraq (Boehmer 1973). Inasmuch as it was
assumed in 1912 (and continued to be assumed for decades) that Urartian
territory encompassed only the area around Lake Van in eastern Turkey,
Thureau-Dangin (1912:ix and map; here fig. 1) reconstructed Sargons route
from Iran to Urartu as involving a passage around the east and north shores
of Lake Urmia in nw Iran, a turning west across the Zagros to Lake Van, and
a circling of that lake before the homeward journey.
The discovery in 1968 of the Urartian site of Qalatgah with its late 9th-
century Urartian inscription (Muscarella 1971) and the extensive explora-
tions and surveys in Azerbaijan conducted by W. Kleiss of the German
Archaeological Institute1 made manifest the assumption that Urartian ter-
ritory from the late 9th to the 7th century bc in fact also included the area
around the western and southern shores of Lake Urmia. This archaeologi-
cally-documented evidence of the extent of Urartian-controlled land has
forced the conclusion that Thureau-Dangins reconstruction itself required
1 The results of these surveys and explorations have been published by W. Kleiss and his
colleagues in AMIran beginning in 1970 and continuing into the 1980s. Kleisss surveys are a
significant contribution to Urartian and Iranian archaeology. See also Kleiss (1976) and Kroll
(1976).
372 chapter eleven
Figure 1. The route and topography of Sargon II

as proposed by F. Thureau-Dangin (1912).
revision.2 Sargons Urartian campaign and his reference to the sea could
only signify that the Assyrian text was concerned with the Urartian area
around Lake Urmia, that Sargon never campaigned in Urartu proper, and
that he never went to Lake Van. What is more, there is no indication from
the text that could allow one to conclude that Sargon circled around Lake
Urmia from the east. Nor is there any archaeological evidence for an eastern
route; in fact, there are no Urartian settlements along the eastern side of
the lake (Kleiss 1976: pp. 1, 2, with map).3 Sargon traveled to Lake Urmia and
2 For the relevant literature offering interpretations of the eighth campaign, see Ziman-
sky (1985: 112, note 55); Lehmann-Haupt (1926: 309); Oppenheim (1960); Kleiss (19691970);
Muscarella (1971); and Reade (1978: 141).
3 Doubts about Sargons alleged circling of Lake Urmia first surfaced in my mind in 1960
when I traveled by truck along its eastern shore from Tabriz to Hasanlu: the lake was too large
moved some distance up the western shore before turning west and south
for the journey to Musasir.
As this investigation is primarily concerned with the identification of
Qalatgah, it will be limited to an examination of the areas cited by Sargon as
being in or bordered by Urartian-controlled territory. The text records that
after leaving Uishdish, apparently a Mannaean-controlled region, Sargon
passed through Subi, Sangibutu (with many cities), Armarili, Aiadi on the
sea (i.e. Lake Urmia), then on to Uaiais, and finally to Nairi-Hubushkia,
from which area he began the journey home.4 Attention will be focused first
on Uaiais, appropriately in the context of its identification with Qalatgah by
Zimansky.
Thureau-Dangin (1912: x, note 1) was the first to suggest that Sargons
Uaiais (Assyrian text, line 298) is a variant of Uisi, Uise, etc., known to
cuneiformists from a number of Assyrian documents, and his observation
has been accepted by all scholars. Believing that Sargon circled Lake Van,
Thureau-Dangin placed Uise (as I shall continue to cite it) at Bitlis, near
the sw shore of that lake; others who accepted the encirclement of Lake
Van agreed with Thureau-Dangin (e.g., Piotrovskii 1959: 105, with map; Bur-
ney 1972: 156). Lehmann-Haupt (1916: 143144; 1926: 310, 317319, 322), who
believed Sargon traveled close to but did not encircle Lake Van, situated Uise
further east, at the modern town of Bashkale; Knig (1955: 68, note 1, 207)
placed it somewhere sw of the lake. The placement of Uise in se Turkey
continued to be accepted in more recent times. ilingiroglu (19761977: 265,
fig. 1) accepted the fact that Uise is near Bashkale; while Levine (1977: 145,
147) and Mayer (1980: 15, fig. 1, 29) situated it further south between the
Upper Zab River, near the modern town of Hakkiri-lemerik, and Lake
Urmia.
(some 80 miles in length) not to have been mentioned if circled. These thoughts reached
fruition in 1968 when I discovered Qalatgah and its inscription.
At least three scholars in recent times continue to believe that Sargon circled the lake from
the east, even though they reject the long route to Lake Van (i.e. a medium route): Wfler
(1976: 20); ilingiroglu (19761977: 259265); Reade (1978: 140, fig. 2, 141). And, to mention
only post-1970 writers, Burney (1972: 138139, 155156), van Loon (1975: 206207), and Kleiss
(1977a: 138140) continue to support Thureau-Dangins long route around the east shore of
Lake Urmia to Van. The short route, essentially to the sw shore of Lake Urmia, first proposed
by the writer in 1971, has been accepted by Levine (1977: 144148, fig. 1), Mayer (1980: 224
225), and Salvini (1982: 387; 1984: 15, 4651). Zimansky (1985: 40, 93, 113, note 81) is ambiguous;
on page 93, however, he summarizes his important research to conclude that the lengthier
routes that have been proposed may be ruled out.
4 For the sequence, see Levine (1977: 137148, and fig. 1 [here fig. 2]), and Zimansky (1985:
41, fig. 7) for a schematic drawing of the route in the northern areas.
374 chapter eleven
To my knowledge it was J.V. Kinnier-Wilson (1962: 109110) who first per-

ceived Uise to lie further east, in nw Iran, presumably positing its identifi-
cation on phonemic grounds with modern Ushnu. Earlier, Lehmann-Haupt
(1926: 341) had connected the city Uisini mentioned in Urartian texts (see
my note 17 below) with Ushnu, but apparently he did not refer to Uaiais/Uise
of the Sargon text. Van Loon (1975: 206207) subsequently made the same
identification without reference to Lehmann-Haupt or Kinnier-Wilson. Not-
ing that the inscription discovered at Qalatgah mentions the city Uise (URU
U-i-se),5 he wondered whether Uise would not be the same as modern
Ushnu. Apparently these scholars were referring specifically to the modern
town of Ushnu (or Ushnuviyeh), 13km west of Qalatgah in the same valley,
and not to the valley at large (see also Reade [1978: 140, fig. 2]). Considering
the Uise-Ushnu identification to be uncertain, M. Salvini (1979: 174) posited
another identification: Es kann aber auch mit der Mglichkeit gerechnet
werden, dass in Uise der alter Name von Qalatgah selbst vorliegt. (I shall
return again to Uise in the Qalatgah inscription.) P. Zimansky (1985: 66
67, 112, note 64), as already noted, came to the same conclusion, apparently
independently of Salvini. He offered as his reasons for such an identification
the Qalatgah inscription reference, the topography of the site, the fact that
it has a rear side (a feature mentioned by Sargon at Uise, see below) and its
geographical position on the lower border of Urartu and in proximity to the
Kel-i Shin Pass, by which Zimansky believes Sargon went to Musasir.
Reinterpretation of the Evidence
I was tempted to accept the equation between Uise and Qalatgah, because
for some time I was uncertain whether Sargon had traveled far up the
western shore of Lake Urmia. If he did not travel up the shore Uise would
have to lie, on the basis of the campaigns sequence, somewhere by or below
its south shore. But a careful reading of Sargons reference to Uise left me
unsatisfied with Salvinis and Zimanskys opinion that Qalatgah was being
discussed. True, Qalatgah may be a great fortress and is situated in a fertile
area, necessary features for the identification. But they are not sufficient
features alone, and no obvious element or characteristic compels one to
5 The Qalatgah inscription is written on a stone block, 0.88 0.36 0.44cm, which is
about half its original length. Fortunately, the opening lines are among those surviving. They
mention that the inscription was written by Ishpuini and his son Menua, which places it in
the last decade of the 9th century bc. For photographs and a drawing of the stone see van
Loon (1975: figs. 13).
connect the site with the textual reference. For example, it is by no means
clear what part of the steep site would qualify as its rear. And regarding the
Uise citation in the inscription, it is unclear from the incomplete context
(see note 5) whether or not it actually refers to Qalatgah itself.
Further study of Sargons text in a search for other possible candidates
that might relate to Qalatgah brought me first to the passages concerned
with the cities of Sangibutu, and particularly the city of Ulhu. And here I
immediately sensed that I was reading a description of Qalatgah.
Ulhu, like Uise and many other cities and districts mentioned by Sargon,
has been moved by modern scholars back and forth across the ancient
landscape, and for the same reasons. Ulhu is mentioned in Sargons text
as a city in Sangibutu, at a distance from other local cities and therefore
apparently in its northern section. Thureau-Dangin (1912: viii), Lehmann-
Haupt (1916: 145, note 1; 1926: 319), Piotrovskii (1959: 105, with map), and, in
more recent times, van Loon (1966: 18) and Reade (1978: 140, fig. 2) situated
Ulhu in the fertile Marand plain, ne of Lake Urmia. Burney (1972: 140, 154)
and Kleiss (19691970: 131132, fig. 3; 1977a: 140) believe it is to be situated
either near Marand or further sw, near the modern town of Shapur, where
earlier Wright (1943: 185) and Laesse (1951: 21, note 2) had placed it. Kleiss
and Burney also suggest that if Ulhu was in the Shapur plain, it is to be
identified with the ancient site of Haftevan Tepe (Kleiss 1976: maps 1 and
2, site no. 32).
As all these identifications were based on the assumption that Sargon
passed around Lake Urmia from the east, thence either west to Van or
southward, they can readily be dismissed. And even if one could assume
that Sargon marched up the west shore of Lake Urmia as far north as
the Shapur Plain, Haftevan Tepe cannot by any archaeological reasoning
qualify as a major Urartian site, as Zimansky (1985: 39, 45, 113, note 100) has
demonstrated.
Ulhu is described by Sargon as a strong city (al dannuti, line 200) situ-
ated at the foot of Mt. Kishpal,6 adjacent to a fertile area with orchards and
gardens. These features do not conflict with the topography at Qalatgah, a
walled city built on the slopes of a mountain and adjacent to a higher one,
and situated in one of the most fertile valleys in Iran.7 What is significant,
6 In the text, the line mentioning the name of the mountain is broken away, but in a
parallel passage in Sargons annals it is mentioned (Thureau-Dangin 1912: 71, line 114).
7 For a description, with plans and photographs of the site, which was extensively sur-
veyed but never excavated, see Kleiss (1971: 6364, figs. 13, 14; 1977b: 7072); and Muscarella
(1971: 42).
376 chapter eleven
however, in Sargons description and what compels attention immediately,

is the reference in the broken lines 202204 to flowing water, and the Assyr-
ian writers positive reaction to them. Line 202 seems to indicate that Rusa
showed/revealed [the place] where the waters gushed forth (Luckenbill
1927: 86; or, hatte das Her]vorkommen von Wasser gezeigt in Meissner
1922: 115). Laesse (1951: 2324, 27) translates the phrase ([mu]-si-e ma-a-
mi) as referring to a water outlet (Rusa revealed the water outlet), and
he notes its parallel use in connection with the earlier description by Shal-
manesser III of the Tigris tunnel, the cave outlet for the river.
Line 203 mentions the building by Rusa of a canal (palgu) that carried the
water, which is said to have flowed in great quantity, like the Euphrates.
Line 204 records other activities of Rusa, namely that countless ditches (or
channels) he led off from its interior (su-ru-us-sa) to be used for irrigation.

The word surru translated here by Laesse as interior is an exact meaning,

as he makes clear (1951: 26, note 36). Thureau-Dangin (1912: 33) also gave
this meaning, but he added an interpretation to the possessive its as
follows, Du sein de ce (canal) il fit Laesse (1951: 2627), however,
cogently noted that the water was not led from the canal because the
its in the Assyrian suffix (sa) is feminine while the previously mentioned
canal (palgu) in line 203 is masculine. But because Laesse believes qanats
are being discussed in this passage, he restores a feminine noun hirtu,
also a canal or ditch, but here translated as qanat, and renders line 204 to
read that Rusa had ditches led out from its (= qanats) interior.8 Moreover,
the insertions of canal and qanat into line 204 are interpretations, not
translations, and, I submit, they mask the obvious intention of the Assyrian
writer to describe the water source at Ulhu.9
8 Qanats are underground tunnels used to collect ground water; the tunnels eventually
reach open air where the collected water flows into irrigation channels. In line 211 of the text,
hirtu is used (Meissner 1922: 116), alongside of which Rusa built a palace. But would Rusa
have built a palace by a qanat, rather than by a canal, whose flowing water could be utilized?
For a lack of qanats in Urartian territory, see Zimansky (1985: 119, note 128).
9 See also the Chicago Assyrian Dictionary 16 (1962: 259): surru, heart, center (of an

object), but where the author also supplies canal: he had countless irrigation ditches flow
out of it (the canal). Note also that Laesse (1951: 27, note 46) and Oppenheim (1960: 142)
both interpret the broken line 201 (They [the people of Ulhu] did [not] drink, they did not
satiate themselves) to signify that water was in short supply at Ulhu until Rusa discovered
new water sources. Line 201 is far from clear, but one notes in the previous preserved phrase in
the same line, his people like fish , which phrase could hardly suggest that Ulhu had no
water until Rusa arrived. What Sargon is clearly noting is the irrigation works built by Rusa,
and not that there was no water at Ulhu until Rusa discovered it.
For where indeed did the water that flowed as copiously as the Euphrates
into the canal have its source but from the interiorfrom the interior of the
city of Ulhu? This meaning surfaced in my original reading of the text and
it was reinforced by the observations given above. Further reinforcement
came when I learned that the word for city (alu) is feminine (personal
communication from L.D. Levine). This fact might allow one to suggest
that the feminine its refers to the city (line 200), and there is no other
antecedent feminine reference in the surviving text.
This vivid description of water gushing from the city of Ulhu precisely
records the nature of the water source at Qalatgah. And, as written in Sar-
gons text, it is a fine example of Oppenheims (1960: 134) astute comment
that the Assyrian writer had in mind a literary concept that sought to cre-
ate a visible perception or imagery, a representation of reality as in the
[carved stone] reliefs. Qalatgah has within its perimeters two separate water
sources: springs that gush forth from the mountain rock; one is at the south,
the other to the west. To this day, its waters are directed into channels used
for irrigation, and there is a pool with a dam that collects this water before
it flows further. It was among the stones of the dam that the inscription was
discovered.10 Nowhere else in the text does one encounter the enthusiasm
expended on the water source of a city, and it is worthy of note that no other
site in nw Iran but Qalatgah qualifies for such enthusiasm.
There is yet another specific observation in the text that catches our
attention, for it too relates to a feature at Qalatgah. As Sargon departs from
Ulhu, and before he reaches Armarili, he mentions Urartian watchtowers
on the summit of mountains (lines 249250). A tower-like structure that
has been interpreted as a watchtower or observation post (or possibly, of
course, a fire-signal tower) was discovered by S. Kroll and W. Kleiss on a
height separated by a saddle from Qalatgahs peak.11 The correspondence of
10 See Kleiss (1971: 64, fig. 14) for the springs and dam (Teich), and Muscarella (1971: 47). No
attempt has been made to date the pool beyond its present usage. The water from the springs
flows in quantity and is most delicious.
11 Kleiss (1977b: 7072, figs. 22, 23), but with no mention of the Sargon reference. Zimansky
(1985: 45) mentions another watchtower near Verachram, south of Lake Sevan.
Sargon also mentions plane trees at Ulhu. Wright (1943: 185, note 60) says they grow
only in Salmas and Urmia (Rezaiyeh), not further north, but he says nothing of the south.
H. Bobeck, however (1968: 286), says they occur in the Zagros from Turkey south to Fars,
growing only in ravines. None occur today at Qalatgah, but (although I do not remember
myself) I am informed by a colleague that she saw them growing in the town of Ushnu. They
are a favorite tree in modern Iran and are planted wherever there is sufficient water, as in
Teheran.
378 chapter eleven
a distinctive architectural structure on a mountaintop in Sargons text and

an archaeological reality may not be ignored.
The chronology of Qalatgahs floruit presents no problem for its identifi-
cation with Ulhu. The site was founded by Ishpuini and his son Menua in
the late 9th century bc, and it continued to function into the 7th century,12
thus encompassing the time of Sargons journey in 714bc.
A final issue to be confronted is the viability of Ulhus location in the
Ushnu valley within the context of Sargons itinerary, with concern here
solely for Sargons position both before and after reaching Ulhu. When
Sargon left the plain of Subi, which was at least on the border of Urartian
territory, if not necessarily within it (Assyrian text, lines 170, 172; Levine 1977:
142), he moved to Sangibutu, whose cities included Ulhu, and thence to
Armarili. If Qalatgah is Ulhu, it follows that Sargon arrived there directly
from the south or the se, which in general satisfies both Levines (1977: 142,
147, fig. 1: here fig. 2) and Mayers (1980: 15, fig. 1) s-n route. Placing Ulhu
at Qalatgah, however, demands a minor adjustment in their positioning of
Sangibutu itself: its northern border must be shifted further north (fig. 3).
One need not be involved here in the location of Subi, except to note that it
probably lay either south of Sangibutu where Mayer placed it or further east,
perhaps but not certainly, near Mianduab as Levine placed it (see note 19
below).
Both Levine and Mayer conclude that Armarili lay directly north of
Sangibutu and adjacent to the south shore of Lake Urmia, and they believe
that Armarili encompassed both the Ushnu and Solduz valleys. Salvini (1982:
387, fig. 2; also 1984: 31, 47, fig. 2) agrees, for he placed Armarili in the Godar
ay valley (but which, judging from his map, seems to involve only the
Ushnu valley, not the Solduz).13 How firmly these scholars hold the view
that Armarili must include the Ushnu valley is not revealed, but if Qalatgah
is Ulhu, then the Ushnu valley cannot have been a region of Armarili. This
leaves the large and fertile valley of Solduz to its east as a claimant (following
12 The chronology is known from the results of many surveys (Muscarella 1971: 4647;
Kroll 1976: 9293, 157159, 169; 1977: 109).

13 Salvini is of the opinion that in the 9th century Hasanlu (Period IV) and Tashtepe
collectively make up Meshta, mentioned only in late 9th- and early 8th-century bc Urartian
texts at Tashtepe and Karagnduz (1979: 177; 1982: 391392; also 1984: 1921, 3233, 47, fig. 2).
He further suggests that the latter inscription actually records the destruction of Hasanlu IVB.
He does not suggest an historical name for Hasanlu-Solduz in 714bc, except to note that the
Assyrians could have employed a different name than Meshta, i.e. Gilzanu (Salvini 1982: 392;
also 1984: 21), as Reade (1978: 140, fig. 2; 1979: 175179) suggested.
Figure 2. The route and topography of

Sargon II as proposed by L.D. Levine (1977).
380 chapter eleven
Figure 3. The revised route and topography

of Sargon II. Drawn by Elizabeth Simpson.
Levine in part). That Sargon does not mention Lake Urmia in connection
with Armarili may cause concern, inasmuch as Solduz at its north reaches
the lake. The omission, however, may merely signify that his troops did not
journey there through the marshes while in Armarili, whereas when in Aiadi
(see below) they could not avoid mentioning the lake. (The western sector
of Armarili would be where Hasanlu is in fig. 3.)
Yet another problem in connection with the ancient nomenclature and
relationship of the Ushnu and Solduz valleys is raised by the fact that after
departing from Ulhu and before arriving at Armarili, Sargon attacked 21
strongholds in a mountainous and fertile region. If the Solduz valley was
Armarili, then those strongholds must have been in the Ushnu valley. Such
an interpretation is not impossible, as the valley is bordered on three sides
by mountains; we may also recall that it was at this stage of his campaign that
Sargon mentioned the watchtowers (see above). What is more, Zimanskys
(1985: 42) reading of this event is that the 21 cities were clearly grouped
together on a single mountain, a view that does not conflict with the
suggestion that the action described occurred near Ulhu, and in the Ushnu
valley. To be fair to the limited evidence, it is not impossible (although less

probable) that these strongholds, or some of them, may have extended into
the nw section of Solduz. If such were the case, it would suggest, assuming
the accuracy of the Assyrians knowledge of Ushnu-Solduz borders, that
Armarili may have been located solely in the eastern section of Solduz,
perhaps extending to or near the Tashtepe area (fig. 3).14 The boundaries of
Armarili are not yet firmly apprehended.
I am well aware that there are no facile solutions to the vertiginous
problems concerned with the accurate matching of ancient toponyms and
modern maps, or with the manner in which the Assyrians defined borders in
714bc (which surely were not always straight lines). Nevertheless, the main
issue I argue herehowever the above adjustments to the ancient map
are receivedis that inasmuch as several scholars consider Sangibutu to
be south of Armarili, and Armarili to have been quite close to Lake Urmia,
Ulhu cannot have been far away from the latter. Collectively, the evidence
demonstrates that equating Ulhu with the site of Qalatgah does not do
injustice to an interpretation of Sargons text.
If Qalatgah is Ulhu, where then is Uise? With some assurance it may be
concluded that if Qalatgah is Ulhu, Uise lay somewhere to its north. All the
internal evidence of the description of the campaign after Sargon departed
Armarili up to the Nairi-Hubushkia visit and the final journey homeward
indicates that he continued to travel north, up the western shore of Lake
Urmia, as Levine was the first to articulate; the lack of Urartian sites on the
eastern shore of the lake and to its east reinforces this conclusion.
Leaving Armarili, Sargon crossed over or passed a cypress mountain,
which probably lay near the modern village of Heyderabad at the sw corner
of Lake Urmia, to reach Aiadi. As all the cities of Aiadi are said to exist on the
shores of the sea (the only time that Lake Urmia is mentioned in the text), it
must be situated not far north of the Solduz valley, most probably along the
narrow shore zone and mountain range to its nw. Sargon then passed over
three rivers to reach Uise,15 rivers that one encounters on the modern map
just north of the modern town of Rasakan. How far north Aiadi extended
north of Heyderabad is uncertain, for we do not know which of the rivers
14 On his 1977 map (here fig. 2), Levine clearly does not include the Tashtepe area in
Armarili. He also (1977: 150) considers the fortress city of Hasanlu (Period III) in western
Solduz to be a city within Armarili. But was it one of the 21 cities of Sangibutu captured by
Sargonor one of the cities of Armarili?
15 Note that in the Ushnu valley there is but one river, the Godar ay, ca. 2.5km south of
Qalatgah.
382 chapter eleven
is to be counted as the first of the three crossed by Sargon. The first river
above Rasakan is the Baranduz ay, running e-w across the plain; north of
it is the Berdesur ay, also running e-w; above it is the Sahr ay and the
Ruzeh ay, both meeting at Urmia-Rezaiyeh, the modern city; still further
north is the Nazlu ay. Whichever river is counted as the first mentioned
by Sargon, however, it appears likely that Uise lay north of the Berdesur
ay.16 Here we leave the modern map and turn to the archaeological map
filled in by the important work of W. Kleiss. From this map it is immediately
clear that only one major site exists in the Urmia plain, and only one
qualifies for consideration as reliably reflecting Sargons description of Uise,
the strongest of all the fortresses of Rusa. That site is Qaleh Ismael Aga on
the south bank of the Nazlu ay.
Sargons text does not expend much space describing Uise except to
record that it was a great fortress, which was stronger than any other of his
[Rusas] fortresses, that it had orchards and forests, and that he captured
it from the rear. Kleiss survey and plan (1977b: 64, figs. 1416) show the site
to be situated in a strategisch sehr gnstiger Position auf einem isoliert vor
dem Gebirgsrand sich erhebenden, doppelgipfligen Felsmassif [which]
liegt an einem alten und auch von dem Urartern benutzten Weg Striking
is the cliff face below the site at its north, directly above the Nazlu ay,
which could be perceived as the rear or back of the site. Qaleh Ismael Aga
is the third largest Urartian site in Iran (Qalatgah has yet to be measured),
a complex site [that] dwarfs all other ruins of the Urartian period in the
plain of Urmiyeh (Zimansky 1985: 39, table 2). The pottery collected in
surveys (Kroll 1977: 98106) consists of Iron Age grey wares (probably all
of Iron Age II) and Iron Age III types. Qaleh Ismael Aga was a powerful
fortress in the late 8th century bc (Zimansky 1985: 10);17 and its geographical
16 The point is, we do not know whether Aiadi extended from, say, the area only between
Heyderabad and the Rasakan plain, consisting of a narrow shore and the adjacent mountains,
or whether it extended further north to the Urmia plain. In the text it is not made clear
whether the three rivers were considered to belong to Aiadi or to Uise.
17 Kroll (1977: 107108) assumes from the non-diagnostic greyware asemblage that the site
was founded by Sarduri or Ishpuini and that the site is older than Qalatgah, even as old as
Tushpa; he also believes that Qaleh Ismael Aga is located in Gilzanu (see note 13 above). A city
named Uisini is mentioned in a text of Ishpuini (Meher Kapisi) and also in a text of Menua
(Akdaman) (Knig 1955: 52, 55, 6768 and notes 10, 31). If Uisini is the Urartian for Uise, as
Knig posits (and also Kinnier-Wilson 1962: 109110), then there was a city Uise existing in
the 9th century bc Sargon (line 285), however, mentions a city in Aiadi called the old Uise
on the shore of Lake Urmia, which is clearly distinct from the city of Uise mentioned after
leaving Aiadi and crossing three rivers. Therefore, it is not clear whether the Urartian texts
position coincides with that of Uise on the lower border of Urartu and on
the Nairi frontier (line 298).
If Qaleh Ismael Aga is Uise, then Levines (fig. 2) and Mayers maps must
be considerably adjusted: Aiadi should be situated along the southern, lower
third, of Lake Urmias west shore and the district of Uise must be shifted
east (i.e. in general, to a position on their maps where Aiadi is written),
presumably to include a large part of the Urmia plain. Aiadi and Uise were
provinces or districts within the borders of modern Iran (fig. 3).
At this stage in the study I learned that an Italian team had excavated
at the site of Qaleh Ismael Aga and published the results in Tra los Zagros e
lUrmia.18 But more than publishing the excavation results (Salvini 1984: 215
239), the identification of the site is discussed by M. Salvini, who concludes
that it was ancient Uise. Before commenting on this identification, however,
he returns (1984: 12, 24) after more study (see Salvini 1979) to the presence
of U-i-se in the inscription from Qalatgah, which site he had previously
tentatively identified as Uise. He is now able to report that the reading of the
name is in fact far from certain, that a lacuna between the allegedly written
i and se could allow for a totally different reading, and his conclusion is that
a new collation is warranted. He further notes that even if the originally
proposed reading holds up, we still do not possess the missing part of the
text that would explain why and in what context the city is mentioned. In
short, the Qalatgah inscription itself cannot at present be used to identify
that site.
With regard to Qaleh Ismael Agas identification with Uise, Salvini (1984:
4651) notes that on the basis of Assyrian documents other than the eighth
campaign, Uise must lie se of Urartu proper and near Hubushkia, and that
the three rivers Sargon crossed fit the local topography of the Urmia plain.
refer to old Uise or to (new) Uise, which leaves Krolls early dating of Qaleh Ismael Aga far
from certain. Old Uise and Uise were not far apart geographically, as Zimansky (1985: 119,
note 130), contra van Loon, correctly noted.
18 I learned this from L.D. Levine, whom I called to share my views on the Ulhu and Uise
identifications. The volume was in my library for some months waiting to be read. It is a rich
book, filled with valuable information and insights on le province orientali dellUrartu. The
section on Qaleh Ismael Aga is only one of many topics discussed. Further, I then wrote to
M. Salvini (January 10, 1986) to share my views on Ulhu and Uise and to call attention to the
fact that on independent grounds we had both identified Uise. On February 3, 1986, by which
time this manuscript was finished and about to be sent off for publication, I received from
Salvini the article he co-authored with P.E. Pecorella (1982) which I did not possess. Here, the
authors already concluded that Qaleh Ismael Aga was to be identified with Uise. In addition,
they cover much of the ground dealt with in Salvini (1979, 1982, and 1984).
384 chapter eleven
He calls attention to the large size of the site and the fact that no other large
one exists in the area, that it protects a main road, and that it has a back
the lower castle over the cliff (see Salvini 1984: 219, 228, figs. 40, 49). And as
a result of his identification of Uise, Salvini (1984: 47, fig. 2) makes the same
adjustments to Levines map as I suggested above for Aiadi and Uise.19
Two independent approaches, one starting from the identification of
Ulhu and tracking Sargons route from that posited position, the other work-
ing from local geographical and topographical observations, have reached
the same conclusion that Qaleh Ismael Aga was ancient Uise. Ulhu and Uise,
it is suggested, are to be considered (and reviewed) by scholars as points
fixed on the ancient map of nw Iran, and they join the two others hitherto
recognized in the general area, Musasir and Tashtepe-Meshta.
Having identified and reunited Ulhu and Uise with their ancient mounds,
the aim of this paper is accomplished. Yet, having tracked Sargon to Uise-
Qaleh Ismael Aga, one hesitates to abandon him. This is especially so,
because my understanding of Sargons text is that his journey from Uise
may have been shorter than has previously been recognized. Leaving Uise,
Sargon went to Nairi where he was visited by the king of Nairi-Hubushkia,
Ianzu, who arrived from his royal city, also called Hubushkia, a distance of
four bru (double-hour). Abruptly, in the next line, and with no topograph-
ical commentary, Sargon is in the royal city receiving tribute. The state of
Hubushkia is considered to be situated in se Turkey, near the headwaters of
the Upper Zab, in the vicinity of lemerik-Hakkiri and Yksekova (Levine
1977: 144; Salvini 1982: 386; 1984: 40, 49). But did Sargon journey far west,
to or near the headwaters of the Upper Zab, some 5060 miles as the crow
flies, without comment, to reach the royal city? I doubt it. Further, after leav-
ing Hubushkia for Musasir, the only recognizable geographical reference
mentioned is the Upper Zab, a river locally called Elamunia, which Sargon
crossed shortly after he started on his commando raid. This reference cannot
be interpreted with certainty to signify that Sargon traveled the long dis-
tance to the Upper Zab headwaters from Uise.
19 Salvini does not refer to Ulhu and has not attempted to identify Qalatgah with an
ancient name, except to place it within Armarili. On his map, figure 2, Armarili seems to
occupy more territory than suggested in his text (Godar ay valley). On this map, Salvini
places Sangibutu south of Mahabad. I do not claim to know how far south on the map
this province is to be placed, but inasmuch as I have argued that the Ushnu valley was the
northern part of Sangibutu, then it follows that I would prefer Subi to be placed in or near
the position where Salvini has written Sangibutu, and that the latter be shifted north.
There are other interpretations that suggest themselves for a reconstruc-

tion of Sargons route to Musasir. Two come readily to mind. If Sargon
crossed west into Turkey (via the Sero Pass?), he did not move far from the
border to arrive at Ianzus royal city, and he turned south close to that same
border. The Elamunia River he crossed will have been one of the eastern
branches of the Upper Zab, not its headwaters to the west.20 Alternatively,
one could posit that Sargon did not immediately depart from Iranian terri-
tory, that his Nairi-Hubushkia (perhaps to be understood as its easternmost
territories) lie in the area due west of Qaleh Ismael Aga, in the Movana-
Zeyveh valleys and/or its adjacent western mountain range. Sargon could
have easily traversed these valleys south to a point where he crossed a range,
either west into Turkey, or sw into Iraq near the Ulagh Dagh. In either
case, from wherever he made the range crossing he will have encountered a
branch of the Upper Zab (fig. 3).21
In any event, nothing in the text, in its presence or absence, indicates that
Sargon led his troops and baggage far into eastern Turkey at a time when he
was already planning to return home. And if his western trip is shortened,
likewise was his southern trip to Musasir. The royal city of Hubushkia surely
lay close to the modern Turkish-Iranian border, west of Lake Urmia.
20 Lehmann-Haupt (1916: 140; 1926: 313) suggests the Topzawa ay; Boehmer (1979: 124)
suggests the Barasgird River, ca. 20 km north of Musasir. But if the Upper Zab/Elamunia was
crossed early in the journey, as it is presented in the text, then the river would have to be
further north. Reade (1978: 141) claims Sargon crossed the Rowanduz ay, but this river lies
to the south of Musasir (Boehmer 1973: fig. 1); see the following note.
21 The routes I propose here are of course speculative, although, I insist, viable and
attractive; no one knows from a map how Sargon got to Musasir. One wonders whether
Sargon traversed the Kel-i Shin Pass, which is best entered from the Ushnu valley.
The problem is that we do not know whether Sargon returned to the Ushnu valley, or
whether earlier Assyrian kings used this pass, and Sargon claims that his route to Musasir was
unique. Further, the pass was controlled by the Urartians (Salvini 1982: 389390; 1984: 43), and
perhaps Sargon would therefore not have used it for this reason and because doing so would
have compromised his surprise attack. Lehmann-Haupt (1916: 139140; 1926: 309310, 313,
326327) van Loon (1975: 207), and Zimansky (1985: 112, note 64) believe Sargon took the Kel-i
Shin Pass; Boehmer (1973: 39, note 24; 1979: 124) and Reade (1978: 140141, fig. 2) do not. The
different opinions on this issue are based on views concerning whether Sargon circled Lake
Urmia and concomitantly the locations of Uise and Hubushkia. Reade, for example, believes
with Kinnier-Wilson (1962: 110) in the encirclement of Lake Urmia and that Hubushkia lay sw
of Lake Urmia near Kaneh. In this view, Sargon approached Musasir from the south.
386 chapter eleven
References
Bobek, H.
1968 Vegetation, in The Cambridge History of Iran. Cambridge: The University
Press: 280293.
Boehmer, Rainer Michael

1973 Zur Lage von Musasir, Baghdader Mitteilungen 6: 3141.
1979 Zur Lage der Stadt Musasir, in VIII Trk Tarih Kongresi, Trk Tarih Kurumu
Basimevi: 123128.
Burney, Charles A.
1972 The Peoples of the Hills. New York: Praeger.
ilingiroglu, Altan A.
19761977 The Eighth Campaign of Sargon II, Jahrbuch fr kleinasiatische For-
schung 4/5: 252269.
Kinnier-Wilson, J.V.
1962 The Kurbail Statue of Shalmaneser III, Iraq 24: 90115.
Kleiss, Wolfram
19691970 Zur Ausbreitung Urartus nach Osten, Istanbuler Mitteilungen 19/20:
125136.
1971 Bericht ber Erkundungsfahrten in Iran im Jahre 1970, Archaeologische
Mitteilungen aus Iran 4: 51111.
1976 Topographische Karte von Urartu. Archaeologische Mitteilungen aus Iran,
Ergnzungsband 3.
1977a Alte Wege in West-Iran, Archaeologische Mitteilungen aus Iran 10: 137151.
1977b Urartische Pltze in Iran, Archaeologische Mitteilungen aus Iran 10: 53
83.
Knig, Friedrich Wilhem

1955 Handbuch der chaldischen Inschriften. Archiv fr Orientforschung, Beiheft 8.
Kroll, Stephan
1976 Keramik urartischen Festungen in Iran. Archaeologische Mitteilungen aus
Iran, Ergnzungsband 2.
1977 Die Oberflchenfunde des Urartu-Surveys 1976, Archaeologische Mittei-
lungen aus Iran 10: 83118.
Laesse, Jrgen
1951 The Irrigation System at Ulhu, 8th Century B.C., Journal of Cuneiform
Studies 5: 2132.
Lehmann-Haupt, C.F.
1916 Musasir und der achte Feldzug Sargons II (714 v. Chr.), Mitteilungen der
vorderasiatisch-aegyptischen Gesellschaft 21: 119151.
1926 Armenien Einst und Jetz, 2, 1. Berlin and Leipzig: B. Behr/F. Feddersen.
Levine, Louis D.
1977 Sargons Eighth Campaign, in L.D. Levine and T. Cuyler Young, Jr., eds.,
Mountains and Lowlands. Malibu, CA: Undena Publications, 135151.
Luckenbill, Daniel David

1927 Ancient Records of Assyria and Babylonia II. Chicago: University of Chicago
Press.
Mayer, Walter
1980 Sargons Feldzug gegen Urartu714 v. Chr. Eine militrhistorische Wrdi-
gung, Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 117: 1332.
Meissner, Bruno
1922 Der Eroberung der Stadt Ulhu auf Sargons 8. Feldzug, Zeitschrift fr Assyri-
ologie 34: 113122.

1971 Qalatgah: an Urartian Site in Northwestern Iran, Expedition 13, 3/4: 4449.
Oppenheim, A. Leo
1960 The City of Assur in 714 bc, Journal of Near Eastern Studies 19: 133147.
Pecorella, P.E., and Mirjo Salvini

1982 Researches in the Region Between the Zagros Mountains and Urmia Lake,
Persica 10: 127.
Piotrovskii, Boris B.
1959 Vanskoe Tsarstvo. Moscow: Vostochnoi Literaturi.
Reade, Julian
1978 Kassites and Assyrians in Iran, Iran 16: 137143.
1979 Hasanlu, Gilzanu, and Related Considerations, Archaeologische Mitteilun-
gen aus Iran 12: 175181.
Salvini, Mirjo
1979 Die urartischen Schriftlichen Quellen aus Iranisch-Azerbaidjan , Ar-
chaeologische Mitteilungen aus Iran, Ergnzungsband 6: 170177.
1982 Forschungen in Azerbaidjan: ein Beitrag zur Geschichte Urartus, Archiv
fr Orientforschung, Beihaft 19: 384394.
1984 La Storia della Regione in Epoca Urartea, 951 and I Documenti, 53134
in Paolo Emilio Pecorella and Mirjo Salvini, Tra lo Zagros e lUrmia. Rome:
Edizioni dellAteneo.
388 chapter eleven
Thureau-Dangin, Franois
1912 Une relation de la huitime campagne de Sargon (714 av. J.-C.). Paris: Paul
Geuthner.
van Loon, Maurits

1966 Urartian Art. Istanbul: Nederlands Historisch Archaeologisch Instituut.
1975 The Inscription of Ishpuini and Meinua at Qalatgah, Iran, Journal of Near
Wfler, Markus
1976 Urartu und Assyrien, in Hans-Jrg Kellner, ed., Urartu: ein wiederentdeck-
ter Rivale Assyriens. Munich: Museum fr Vor und Frhgeschichte, 1921.
Weidner, Ernst F.
19371939 Neue Bruchstcke des Berichtes ber Sargons achten Feldzug, Archiv
fr Orientforschung 12: 144148.
Wright, Edwin M.
1943 The Eighth Campaign of Sargon II of Assyria (714BC), Journal of Near
Zimansky, Paul
1985 Ecology and Empire: The Structure of the Urartian State. Chicago: The Ori-
ental Institute of the University of Chicago.
chapter twelve
SURKH DUM AT THE METROPOLITAN MUSEUM OF ART:

A MINI-REPORT*
With a Contribution by Elizabeth Williams-Forte
Abstract: In 1938 Erich Schmidt, taking time out from his major work at Persepolis,
excavated for three weeks the site of Surkh Dum in eastern Luristan, in western
Iran. Although very little has been published on the finds and architecture, aside
from two brief and summary reports by Schmidt and Maurits van Loon, Surkh Dum
is recognized by Iranian archaeologists to be one of the most important sites in
Luristan, and in Iran in general. Not only was Surkh Dum a settlement site, rather
than a cemeterywhich is the typical circumstance in the archaeological history
of Luristanbut many hundreds of objects of bronze, ivory, bone, faience, and
terracotta, as well as about 200 cylinder and stamp seals, were recovered. To date,
only seven of the objects have been published, and nothing has been published
about the two buildings partially uncovered. In 1943 The Metropolitan Museum of
Art acquired 41 objects excavated at Surkh Dum, only five of which had previously
been published. Because of the importance of the material for modern knowledge
of the art and archaeology of Luristan, an area plundered since the late 1920s, and
the source of countless thousands of unexcavated objects, the presentation of even
a small group of excavated artifacts from Luristan is considered to be of great value.
The present paper offers a history of our present knowledge of the site, a tentative
discussion of its chronology, and a catalogue discussion of the Surkh Dum material
in The Metropolitan Museum of Art.
The Background
Following his air reconnaisance trip in 1937 organized to survey sites for
future excavations, Erich Schmidt in the summer of 1938 began his arduous
epic overland journey into Luristan, in western Iran, through the ragged
crests of the Zagros in search of the sites plotted by air. After weeks of
failure he struck his archaeological Bonanza: the site of Surkh Dum in
the Kuh-i Dasht region of eastern Luristan (fig. 1).1 At the time, Schmidts
* This article originally appeared as Surkh Dum at the Metropolitan Museum of Art: A
Mini-Report. Journal of Field Archaeology 8, no. 3 (1981): 327351.

1 I. Erich Schmidt, The Second Holmes Expedition to Luristan, Bulletin of the American
390 chapter twelve
Persepolis expedition staff was on temporary loan to the American Institute

for Iranian Art and Archaeology, the official sponsor of the campaign, under
the title of the Second Holmes Expedition to Luristan.2 Work pressure of
the ongoing Persepolis excavations (completed in 1939) prevented Schmidt
from writing more than a brief report on his Luristan discoveries (which
included other sites besides Surkh Dum); and because of the great length of
time needed to publish the three Persepolis volumesa task that occupied
the remaining years of his lifeSchmidt was never able to publish the Surkh
Dum material in full. All that was given in the 1938 report was a bare listing of
some of the types of objects found and a brief mention of a building, which
on the basis of the plan and the objects recovered indicate that the Surkh
Dum ruin was a sanctuary, a temple of the first half of the first millennium
B.C.. The few photographs published are group shots, and while some show
the juxtaposition of the finds, in only one or two instances is it possible to
recognize a specific artifact (below, No, 5).3 With regard to the building, no
plan was furnished; it was simply revealed that it had a brick superstructure
over a stone foundation, and that it had a terraced square at the center of
the main room which was considered to be an altar.
The types of objects listed consist of a ram-headed stone pestle (Schmidts
fig. 7, A), bronze mirrors, pins, wands, a whetstone, male and female fig-
urines in frit as well as bronze, and over 200 cylinder and stamp seals.
Schmidt also mentioned small bronze plaques, including many fragments
Institute for Iranian Art and Archaeology V, 3 (1938) 205. The name of the site has been
published as Surkh Dum, Dum Surkh, Surkah Dum, Surkh-i Dum, Surkh Dum-i Luri: see Oscar
White Muscarella, Unexcavated Objects and Ancient Near Eastern Art: Addenda (Undena 1979)
12, note 8; Helene Kantor, Embossed Plaques with Animal Designs, JNES V (1946) 234, note 3,
235, 237; Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin 1969) 87,
143145, 150, 159, 169, 188; Pierre Amiet, Les Antiquites du Luristan (Paris 1976) I wish to thank
Trudy S. Kawami for reading my manuscript and for sharing with me her notes and thoughts
on Surkh Dum.
2 The nature of the sponsorship is not made clear in Schmidts report: on page 204 it is
stated that the staff of the Persepolis Expedition was put at the disposition of the American
Institute.
3 Schmidts report appeared very shortly after his Luristan campaign was terminated, and
as stated on page 208: the individual objects have not yet been photographically recorded.
The excavations at Surkh Dum were obviously rushed, taking only three weeks time.
V.E. Crawford informs me that Richard Carl Haines, the architect, told him that the finds
appeared in such profusion that excavation had to cease in order to register the objects; this
may explain why some of the objects do not have a SOR field number painted on them. (SOR
derives from Sorkh Dum, which is a variant spelling of Surkh Dum.) The campaign was also
hampered by the terrain; among other problems, fodder for the donkeys used in transport
had to be imported.
surkh dum at the metropolitan museum of art 391
Figure 1. Map showing Surkh Dum in northern Luristan in

Iran. (The map is based on a map published by Jorgen
Meldgaard in 1963; see below, note 21.) The insert locates
Luristan (framed area) within a larger geographical region.
392 chapter twelve
with scenes in repouss (italics added). Concerning the wands, of which

hundreds were found, it seems that this category included both disc-headed
pins as well as those with heads in the form of figures in the round, and
made of bronze, iron, or bone. Many objects were recovered either enclosed
within (Schmidts figs. 8 and 9), or sticking in the interstices of the walls,
while others were found clustered or scattered in the rooms. And Schmidt
pointedly noted that no horse trappings, one of the most characteristic
artifacts of the Luristan repertory, were recovered. The above summary is
all the first-hand information made available in publication concerning the
threeweek campaign at Surkh Dum.
In the catalogue of an exhibition of Persian art on view in New York City in
1940, P. Ackerman4 briefly referred to the site, mentioning the building and
its various small rooms, as well as the altar. The catalogue also published in
list form a number of objects deriving from Schmidts excavations at Surkh
Dum on loan to the Exhibition from the Oriental Institute of the University
of Chicago, which were juxtaposed to other objects listed as being in dealer,
private, and museum collections.5 Some of the objects listed in the catalogue
had not been mentioned in Schmidts report: bracelets with animal termi-
nals, rings, axes (one spiked, one a miniature; see No. 24, below),6 a buckle,
an anklet, an arrow or javelin point, a frog pendant (or pin?), a bronze vessel,
and pincers, all bronze. Only a year later, in an article co-authored by Ack-
erman and A.U. Pope,7 the date of the discovery of Surkh Dum was given
as 1937, and, puzzlingly, the sanctuary was described as a circular stone
building. Following this erroneous statement, they published a number of
bronzes that implicitly were attributed to clandestine finds at Surkh Dum.8
No matter how one perceives their place of origin, these bronzes did not
derive from Schmidts excavations. Both the date error and the reference
to a, this time, small circular sanctuary were repeated some years later,
4 Phyllis Ackerman, Guide to the Exhibition of Persian Art, 2nd ed. (New York 1940) 541.
5 Ibid. 130, KK, 532, 534536, 540, 542548.
6 Ibid. 547, XX. Both Ackerman, 532, F, and Maurits van Loon in his review of Dark Ages
and Nomads c. 1000 B.C., Machteld Mellink, ed. (Istanbul 1964), in BibO XXIV, 1/2 (1967) 24,
mention a spiked axe with a lion-head juncture. This axe is now in the University Museum
(SOR 1633) and has a blade and spike formation of the same type as P.R.S. Moorey, Catalogue
of the Persian Bronzes in the Ashmolean Museum (Oxford 1971) 5355, no. 20: but with the lions
head facing the spikes, not the blade. Thus, it is not the same form as suggested by Moorey,
ibid, 51, nos. 14, 15, and in his Ancient Persian Bronzes in the Adam Collection (London 1974)
43, nos. 7, 8.
7 Prehistoric Nature Worship in Western Iran, ILN (March I, 1941) 292.
8 See Muscarella 1979, op. cit. (in note 1) 13.
in 1955, by Ackerman.9 Two years later, however, Ackerman more correctly

noted that there were two buildings excavated at Surkh Dum, a Temple and
a small stone sanctuary, but there was no mention of their shape.10 Within
this period at least two other references to Surkh Dum were published,11
both giving information not previously recorded and likewise not revealing
sources. In 1945 Pope referred to a single Assyrian temple plan to describe
the building in which he claimed were found fragments of greenish glazed
tiles, with griffin and sphinx designs. And in 1946 M. Bahrami, apparently
following Pope in part, also cited one building, described as having a main
hall with six entrances as well as a fire altar, and which was decorated with
glazed wall tiles. He also published two disc-headed pins known to him
from the antiquities market but which he claimed derived from Surkh Dum
(infra).
It was not until 1967, 29 years after Schmidts original publication, that
another report on the site was made available, this one apparently based on
Schmidts field notes. In his review of Dark Ages and Nomads, M. van Loon
presented a summary of the various types of objects recovered from the
sanctuary in anticipation of a full account shortly to appear.12 Of immediate
interest is that, whereas Pope in an editorial aside in Schmidts 1938 report13
claimed that no levels could be determined, and consequently there is
no dependable archaeological evidence for dating , van loon claimed
that there were at least three building phases to which he assigned dates
(presumably his, not Schmidts); this chronology he later modified.14 Listed
by levels this time, Van Loon discussed some (but not all) of the objects
9 Phyllis Ackerman, The Gemini are Born, Archaeology 8, 1 (1955) 26; the date 1937 is
repeated.
10 Phyllis Ackerman, A Luristan Illustration of a Sunrise Ceremony, Cincinnati Art Muse-
um Bulletin 5, 2 (1957) 4. Trudy Kawami has shown me a sketch plan of the site: two buildings
were incompletely cleared. The larger has 17 rooms or areas, the less excavated has five; no
circular walls exist. In the larger building there is a big room, not centered, that has a central
unit, the altar, no doubt.
11 A.U. Pope, Masterpieces of Persian Art (New York 1945) 16; Mehdi Bahrami, Some
Objects Recently Discovered in Iran, Bulletin of the Iranian Institute VI, 14 (1946) 71. The
sketch plan mentioned in my note 10 does show six entrances to the large room. Bahrami
may have had access to the same plan in Teheran. We will have to await a final report to get
information about the tiles and to learn if the central unit in the large room was in fact a fire
altar.
12 Van Loon, op. cit. (in note 6) 2324. Van Loon himself has the responsibility for the final
publication.
13 Schmidt, op. cit. (in note 1) 208, note 3.
14 In BibO 29 (1972) 69, note 22.
394 chapter twelve
mentioned by Schmidt, e.g., disc-headed pins, and added others not specifi-
cally mentioned either by Schmidt or Ackerman: a square-framed openwork
plaque (a wand?), lobed bronze rings, bone lion pins on an iron shank,
pottery, incised frit vessels, and, surely of some importance, dedicatory
inscriptions. He also enumerated the types of pins recovered, straight ones
with a variety of decorated heads, and disc-headed, decorated in repouss.
Van Loon specifically noted that the typical Luristan standards were non-
existent at Surkh Dum; and it is relevant to mention at this point that neither
Schmidt not Van Loon mentioned any quivers or large decorated plaques,15
nor objects of precious materials. With regard to the pottery cited, it would
be rash to reach firm conclusions on the basis of the brief descriptions given,
but mention is made both of triangles used as decoration and of the absence
of red and grey wares, indications that we are dealing with a post (or late)
Iron II/Iron III context.
To date, the above cited reports and obiter dicta contain all the basic infor-
mation from first- and second-hand sources available about the excavations
of Schmidt at Surkh Dum. Whether all the types of objects recovered have
been mentioned, or there are still others unreported, will not be known until
a final publication appears (in the meantime, see below).
Objects Allegedly from Surkh Dum
Up to the present time only six of the many hundreds of objects excavated at
Surkh Dum and mentioned by Schmidt, Ackerman, and Van Loon have been
recorded and published with photographs:16 four cylinder seals (Nos. 33
35, 40, below), a frog-headed pin (No. 11 below), and a glazed faience ves-
sel (cf. No. 31, below). The objects listed by Ackerman in the 1940 cata-
logue of the Persian Exhibition were not illustrated. Trudy S. Kawami has
recently informed me, however, that yet another object from Surkh Dum has
been published, albeit without a provenience reference, and it is an impor-
tant object indeed. In his work on decorated bronze nipple beakers, Peter
Calmeyer17 published a drawing of a small fragment of one of these beakers
15 It is not clear what Van Loon means on page 25 by hammered bronzes; and it seems
misleading to claim that they all seem to come from the top level, inasmuch as he placed the
hammered disc-headed pins in the earlier levels. What is more, his dating of these hammered
bronzes to the 7th century is based on an incorrect, low dating of the nipple beakers (see
below).
16 Muscarella, op. cit. (in note 1) 12.
17 Peter Calmeyer, Reliefbronzen in babylonischem Stil (Munich 1973) 3233, A24.
in the Teheran Museum (No. 1124), where it is listed merely as from Luris-
tan. According to Kawami, who saw the original drawing in the Oriental
Institute in 1974, this fragment was found in the refuse dump at Surkh Dum
and has the field number SOR 1712. No doubt the fragment was given to the
Teheran Museum as part of its division of the excavated finds and presum-
ably the records were lost or misplaced so that Calmeyer was not informed
of the provenience. The discovery that a decorated nipple beaker was exca-
vated at Surkh Dum is significant, for, to my knowledge, among the many
scores known to exist, it remains the only one to derive from a controlled
excavation.18 Furthermore, the provenience is a site in Iran, which reinforces
the generally held view that the beakers as a group derive from that area.19
Equally important is the occurrence of a decorated nipple beaker in a sanc-
tuary, a fact that refutes Calmeyers suggestion that all the beakers must have
derived from tombs,20 and which also neatly demonstrates that excavated
material is the only evidence that can be used to form hypotheses regarding
alleged proveniences and the functional value of an artifact.
At the same time, a number of objects, all of which derived from the
antiquities market, that is, from clandestine plundering, have casually been
attributed to Surkh Dum by a number of scholars. In his 1938 report Schmidt
mentioned that digging by local villagers had occurred at Surkh Dum before
he began his campaign and a number of scholars have either called atten-
tion to this fact or have claimed that after Schmidt left more digging took
place.21 Thus, after 1938, when a dealer attributed an object to Surkh Dum
it was found to be convenient to accept the claim, given the references to
the clandestine activitybut which information, let it be recognized, came
from the same dealers.22 It should also be noted that some scholars were mis-
led by information from individuals who supposedly were in a position to
18 Oscar White Muscarella, Decorated Bronze Beakers from Iran, AJA 78 (1974) 243244;
idem, review of Calmeyer 1973, op. cit. (in note 17) in JAOS 97, 1 (1977) 77.
19 For discussion see Muscarella 1974, ibid. 243245, 248249; 1977, ibid. 77.
20 Calmeyer 1973, op. cit. (in note 17) 123, 151, 231, 233, and discussed in Muscarella 1977, op.
cit. (in note 18) 77.

21 Pope 1941, op. cit. (in note 7) 292293; Bahrami, op. cit. (in note 11) 71; Moorey 1971, op.
cit. (in note 6) 1920; idem, Some Elaborately Decorated Bronze Quiver Plaques Made in
Luristan, c. 750650 BC, Iran XIII (1975) 20; Amiet, op. cit. (in note 1) 1; Roman Ghirshman,
Bichapour II (Paris 1956) 120, note 1; idem, The Arts of Ancient Iran (New York 1964) 48; Andr
Godard, L Art de lLran (Paris 1962) 52; Jorgen Meldgaard, et al., Excavations at Tepe Guran,
Luristan, ActaA XXXIV (1963) 98, note 5; D. de Clercq-Fobe, Epingles votives du Luristan
(Teheran 1978) Introduction, page 2.
22 More pins seem to have been offered for sale after 1938 than before.
396 chapter twelve
know what in fact came out of the site.23 Others made what they considered
to be intelligent guesses concerning attribution, on the assumption that cer-
tain types of objects, disc-headed pins for example, derive only from Surkh
Dum. We will now examine a few of these misattributed objects.
The well-known bronze quiver plaque in the Metropolitan Museum of
Art, with its rich and important evidence of style and iconography for west-
ern Iran,24 was purchased from A.U. Pope in 1941, two years before he sold
the Museum the material from Surkh Dum to be discussed below. The site
of Surkh Dum was not mentioned in the transaction and the only informa-
tion given was that the plaque came from the Kuh-i Dasht region in Luristan.
Of significance for the issue of origin is that a year before the purchase Ack-
erman, in the aforementioned Persian Exhibition catalogue,25 published the
plaque, naming as its owner the antiquities dealer R. Rabenou, thus reveal-
ing candidly that the object was available on the art market, and that, con-
sequently, its final ancient resting place cannot be known.26 Yet within a few
years, by 1945, Pope published the quiver as coming from Surkh Dum27 and
from this time onward most scholars who cited the plaque also assigned it
to that site.28 A puzzling exception was Ackerman, who in 1955 repeated her
23 See Muscarella 1979, op. cit. (in note 1) 56, 13; cf. also statements by Ren Dussaud,
Anciens Bronzes du Luristan et Cultes iraniens, Syria XXVI (1949) 213, or Pope apud Schmidt,
op. cit. (in note 1) 210211, note 5.
24 MMA number 41.156; Moorey, 1975 op. cit. (in note 21) 19, note 1, 2426, pl. 1; Edith
Porada, Iranische Kunst, in Der alte Orient, W. Orthmann, ed. (Propylen Kunstgeschichte:
Berlin 1975) 397398, no. 317. From what information I have been able to gather, the quiver
was on the market at least by 1939. Stylistically judged, the quiver is definitely Iranian in
manufacture.
25 Ackerman, op. cit. (in note 4) 115: not 199as in Muscarella 1979, op. cit. (in note 1) 13,
which was a misprint.

26 See Muscarella 1979, op. cit. (in note 1) 13; also Ackerman 1955, op. cit. (in note 9) 27,
and 1957, op. cit. (in note 10) 4. There are no references to quivers in the Surkh Dum field
records: cf. Moorey 1975, op. cit. (in note 21) 19, note 1, where the Surkh Dum archives are
cited and where the distinct impression is given that the quiver was excavated by Schmidt.
V.E. Crawford told me that Carl Haines claimed that no works of art were found at Surkh Dum.
Presumably it could be argued by the interested parties that the quiver, and others, were
found after Schmidt left, by the local peasants: but this type of argument involves gratuitous
guessing, not archaeological reasoning, and begs many questions.
27 Pope 1945, op. cit. (in note 11) pl. 15.
28 Dussaud, op. cit. (in note 23) 213, citing Mme. Godard as an authority; Calmeyer 1969,
op. cit. (in note 1) 87, note 289, 290, D, 145, 159; Amiet, op. cit. (in note 1) 85; Helene Kantor,
The Shoulder Ornament of Near Eastern Lions, JNES VI (1947) 258; Edith Porada, Nomads
and Luristan Bronzes, in Mellink, op. cit. (in note 6) 27, note 61; idem, The Art of Ancient
Iran (New York 1965) 8789, 236, note 10; in her 1975 paper, op. cit. (in note 24) she modified
her position, omitting the site attribution; Henrik Thrane, Archaeological Investigations in
original claim that it was known through the antiquities market.29 At least
three scholars further believed that not only this quiver, but still others, were
found in the Surkh Dum sanctuary, or belonged to a Surkh Dum group.30
Disc-headed pins are perhaps the main type of object (aside from quiv-
ers) most often associated with Surkh Dum, because indeed many were
recovered there. To date, however, not a single example of those excavated
by Schmidt has been published (now see No. 3, below). What has been
published, on the other hand, are scores of examples, all of which derived
from the antiquities market, and the majority of which have been casually
attributed to Surkh Dum. Disc-headed pins began to surface during the time
of the earliest appearance of the Luristan bronzes,31 but whether they were
plundered from one or several sites is of course not known. This latter point
notwithstanding, some scholars in the past have claimed to know the mul-
tiple sources. Thus, Pope, in an editorial comment about the Surkh Dum
wands, stated that closely related pins have been found at other sites ,
while Amiet believes that disc-headed pins have been dcouvertes dans
des tombes.32 Presumably, the pins discussed were those known before
Schmidt went to Surkh Dum, for after 1938 only Surkh Dum is mentioned.
And it was Pope again33 who, forgetting or ignoring his 1938 comment about
other sites, began to cite stray disc-headed pins as coming from Surkh Dum.
As with the quiver, various scholars over the years followed Popes lead
probably because of his institutions role in the excavationsand accepted
as received knowledge that any stray disc-headed pin, including these in the
Coiffard and Graeffe collections, came from one site, from Surkh Dum.34
Western Luristan, ActaA XXXV (1964) 159, note 7 (not 59, as Muscarella 1979, op. cit. [in note 1]
13); Irene Winter, A Decorated Breastplate from Hasanlu, Iran (Philadelphia 1980) 18, note 89.
29 Ackerman 1955, op. cit. (in note 9) 27, 29, fig. 4.
30 Ghirshman 1964, op. cit. (in note 21) 70; Moorey 1975, op. cit. (in note 21) 1926; E.D. Phil-
lips, The People of the Highland, in Vanished Civilizations, E. Bacon, ed. (New York 1963)
227; Amiet, op. cit. (in note 1) 85, claims that the David-Weill quiver came from Surkh Dum
certainement.
31 Viz. A.U. Pope in ILN (September 6, 1931) 390, fig. 16; Andr Godard, Les Bronzes du
Luristan (Paris 1931) pl. XXXIV; Leon Legrain, Luristan Bronzes in the University Museum
(Philadelphia 1934) pl. VI; Ernst Herzfeld, Das Ornament nach der Mitte des II. Jahrtausends,
AMIran VIII (1937) 156157, Abb. 118; J.A.H. Potratz, Scheibenkopfnadeln aus Luristan, AfO
XV (19411945) 39, note 5; cf. Ghirshman 1956, op. cit. (in note 21) 120, note 1, and Clercq-Fobe,
1978 op. cit. (in note 21) Introduction, page 1.
32 Pope apud Schmidt, op. cit. (in note 1) 210, note 5; Amiet, op. cit. (in note 1) 75.
33 Pope 1945, op. cit. (in note 11) 16, pl. 16A, B.
34 Viz. Kantor 1946, op. cit. (in note 1) 234 and note 3; Bahrami, op. cit. (in note 11) 7173,
figs. 1, 2 (fig. 1 = Ghirshman 1964, op. cit. [in note 21] fig. 490, listed as from the Teheran art
398 chapter twelve
Not so incidentally, one of the pins published by Pope35 is gold: Several

gold pieces of high quality were found at Surkh Dum, including a disc.
This pin is the very same one published five years earlier by Ackerman in the
Persian Art Exhibition (pages 134135), where we are informed that Kuh-i
Dasht excavations (sic), conducted by commercial diggers, have also yielded
a very few pieces of gold , the gold disc-headed pin being one of the finds.
But here the pin is listed as being in the possession of an antiquities dealer,
A. Rabenou: an alleged clandestine find of gold in 1940 becomes in 1945 a
find from Surkh Dum. Moreover, as already noted, no gold was excavated at
Surkh Dum. What is more, the pin is not necessarily ancient. These issues
notwithstanding, some scholars have readily accepted the gold pin as both
genuine and as excavated from Surkh Dum.36 Still other objects that I regard
as forgeries have been assigned by some scholars, including Pope, to Surkh
Dum, e.g., a bronze ombos, a silver plaque, and a silver disc-headed pin, all
from the notorious Zurvan group, and a bronze disc and disc-headed pin.37
And a bronze strip, to my mind of suspicious nature (although I am not sure
whether or not it is genuine), was also published by Dussaud as provenant
aussi de Surkh Dum,38 although nothing like it was mentiond by Schmidt or
Van Loon (see also note 54).
Chronology
Although it is to be understood that pending a final publication all com-

ments and conclusions are tentative, because they are based on incomplete
information, a brief note regarding the dating of the Surkh Dum finds is rel-
evant. The discrepancy between Popes and Van Loons claims regarding, on
market); Dussaud, op. cit. (in note 23) 196205, figs. 17, pl. IX, X; Ghirshman 1956, op. cit.
(in note 21) 120122; Calmeyer 1969, op. cit. (in note 1) 87, note 288, 145, note 463: but cf. 143,
note 458; Amiet, op. cit. (in note 1) 1, 75; Clercq-Fobc, op. cit. (in note 21) Introduction, p. 2,
115; Louis vanden Berghe, Archologie de l Iran Ancien (Leiden 1959) 93, 276, pl. 123eh; Pierre
Amandry, Un Motif scythe en Iran et en Grce, JNES XXIV, 3 (1965) 151, with Surkh Dum in
quote marks.
35 Pope 1945, op. cit. (in note 11) 16, pl. 16A.
36 See Muscarella 1979, op. cit. (in note 1) 8, Iranian no. 6.
37 Cited by me as forgeries in Unexcavated Objects and Ancient Near Eastern Art, in
Mountains and Lowlands, eds. L.D. Levine and T.C. Young, Jr. (Undena 1977) 171, 172, nos. 1, 2,
5, 173, note 75; see also Muscarella 1979, op. cit. (in note 1) 3, nos. 3 and 5, also page 6. From
private sources I know that Pope was the vendor of the silver plaque and that he attributed it
to Surkh Dum. I recently examined the pin illustrated in Dussaud, op. cit. (in note 23) fig. 6,
pl. x; cf. Muscarella 1977, ibid. 173, note 75: I have no doubt that it is not ancient.
38 Dussaud, op. cit. (in note 23) 210, fig. 10; Muscarella 1977, op. cit. (in note 37) 175, no. 56.
the one hand, the alleged lack of stratigraphy, and on the other, the existence
of several phases, has already been mentioned. Inasmuch as Van Loon had
access to Schmidts records, we must assume his observations to be correct
and that the existence of floor levels and rebuildings are implied in his state-
ment. Defining these phases, Van Loon claimed that there was a 10th, a 9th,
an 8th, and a 7th century level (four phases?), and that the sanctuary was
in existence from the 10th through the 7th centuries bc, a period of over
300 years.39 Five years later the dates of some of the levels were revised: the
9th century level is actually 8th century; the 8th century level is actually late
8thearly 7th century; nothing was said of the old 10th or 7th century levels.40
Nowhere, in either of the two statements on chronology, is an explanation
given for the determination of dating, although one may infer that it is based
on art historical analyses of the artifacts.
Actually, an important clue concerning dating exists in the presence of
iron, which Van Loon noted occurs in all levels, including the earliest. This
fact alone surely suggests that there is no 10th century level at Surkh Dum,
for there is no evidence to date that the presence of iron is documented
anywhere in Iran before the late 9th century bc.41 It therefore follows that
the earliest level cannot predate the 9th century bc at the earliest, and may
actually not pre-date the 8th.
Confusing the issue of chronology at Surkh Dum is the material of pre-
sumably earlier periods (i.e., earlier than the posited late 9th/8th century
date for the incipient level) recovered there: cylinder seals, a spiked axe with
a lions mask juncture (see note 6), duck-headed pins (see No. 13, below), and
the nipple beaker fragment mentioned above.42 Spiked axes with crescent-
shaped blades are generally dated to the last centuries of the 2nd millen-
nium bc on the basis of related inscribed pieces, none of which, however, has
the zoomorphic juncture,43 and the recent invaluable discoveries of Vanden
Berghe at Bard-i Bal and Kutal-i Gulgul in Luristan have not only furnished
additional Luristan locations for the type (one from the latter site has a
39 Van Loon 1967, op. cit. (in note 6) 24.

40 Van Loon 1972, op. cit. (in note 14) 69, note 22.
41 R. Pleiner, The Beginnings of the Iron Age in Ancient Persia, Annals of the Nprstek
Museum 6 (1969) 34; Louis vanden Berghe, La Chronologie de la Civilisation des Bronzes du
Pusht-i Kuh, Luristan, Proceedings of the 1st Annual Symposium of Archaeological Research
in Iran (Teheran 1973) 4; Vincent Pigott, The Question of the Presence of Iron in the Iron I
Period in Iran, in Mountains and Lowlands, op. cit. (in note 37) 218, 223, 226, 231.
42 Schmidt, op. cit. (in note 1) 208, note 3, 210; see also Elizabeth Williams-Fortes discus-
sion of the seals, infra.

43 Calmeyer 1969, op. cit. (in note 1) 6670; Moorey 1971, op. cit. (in note 6) 4951.
400 chapter twelve
zoomorphic juncture), but also have demonstrated that they continued to

be used until ca. 1,000900 bc.44 From these same sites and from the same
time range Vanden Berghe has also excavated duck-headed pins, animal-
headed bracelets, and a bird pendant, none dissimilar to those from Surkh
Dum (see below, Nos. 13, 19, 23).45 Concerning the dating of the bronze nip-
ple beakers, it seems almost certain that none post-date the 9th century, and
that as a class they were made in the 10th9th centuries bc.46
Were these objects heirlooms dedicated in the sanctuary as precious pos-
sessions, does their presence indicate an earlier date for the sanctuary than
posited here, or does their presence simply indicate a relatively longer life
for these objects than hitherto recognized? The answers are not readily
available and one falls back on expectations of a final report, where one will
be able to see the corpus of material as recovered, rather than viewing iso-
lated objects out of context. While some scholars have followed Van Loons
high dating for the life of the sanctuary, many have nevertheless tended to
support an 8th7th century date for most of the objects recovered.47
Objects from Surkb Dum in

The Metropolitan Museum of Art
In 1943 The Metropolitan Museum of Art acquired by purchase from A.U.

Pope of the American Institute of Iranian Art and Archaeology a group
of 41 objects excavated by Schmidt at Surkh Dum: 24 bronzes, six bone
or ivory objects, fragments of two faience vessels, and nine cylinder seals.
So far as it is possible to learn from published information and private
communication, the finds from the site were divided first in Iran with the
44 Louis vanden Berghe, Luristan Prospections Archologiques dans la Region de Badr,
Archeologia 36 (1970) 10, 13; idem, Recherches Archologiques dans le Pusht-i Kuh Luris-
tan, Bastan Chenasi va Honar-i Iran 6 (1971) 2021, 26, figs. 11, 13, 28; idem, Recherches
Archologiques dans le Luristan, Iranica Antiqua X (1973) 35, fig. 20, pls. XVI, XVIII, 1; idem,
La Ncropole de Kutal-i Gulgul, Archeologia 65 (1973) 18, 22, 24, 25; for revised dating, see
Vanden Berghe 1973, op. cit. (in note 41) 4; Edith Porada, Ancient Persian Bronzes, Apollo
(February 1979) 142.
45 See here notes 88, 94, 106.
46 Calmeyer 1969, op. cit. (in note 1) 224228; Muscarella, 1974 op. cit. (in note 18) 243249;
idem, 1977, op. cit. (in note 18) 77.

47 Moorey 1971, op. cit. (in note 6) 20; idem 1975, op. cit. (in note 21) 19; Pierre Amiet Un
Carquois du Luristan, Syria LI (1974) 244; idem 1976, op. cit. (in note 1) 30, 75, 103. Cf. Clercq-
Fobe, op. cit. (in note 21) Introduction, page 1, who dates the site as beginning in the second
half of the 2nd millennium bc, based on a misunderstanding of Schmidts statements and
Popes comments about early seals at Surkh Dum.
Iran Bastan Museum, and the remainder subsequently in the United States
among the Oriental Institute of the University of Chicago, the University
Museum of the University of Pennsylvania, Mrs. Boyce Thompson, and
Pope. It was the latter collection that eventually came to the Metropolitan
Museum, but it is not known to me whether some pieces in Popes share
were sold or given to other institutions or individuals, or whether his whole
share arrived intact.48
The objects have been in the Metropolitan Museum for almost 40 years,
during which time only four cylinder seals and a frog-headed pin have
been summarily published. Inasmuch as many hundreds of objects were
excavated at Surkh Dum, the Museums collection represents only a fraction
of the total. Yet it may be stated without elaboration that the value for
Iranian archaeology in presenting even this fraction should be obvious.
For, given the recognized significance of the finds from one of the most
important sites in Iran, and one of the few excavated settlements in Luristan,
as well as the attribution to the site of the many dealer-derived objects
mentioned above, a publication is surely long overdue. With these thoughts
in mind, the present paper is to be considered not so much a preliminary,
but rather a mini-report of the Surkh Dum excavations.
In the lists that accompanied the objects sent to the Metropolitan Muse-
um in 1943, the only information supplied was a Surkh Dum field number for
each object, sometimes a plot number and a pr (presumably a plot record)
number, and rarely a R (room) reference; only some of the objects have a
field number painted on it (see Note 3). In a few cases a reference to the
Persian Art Exhibition Catalogue of 1940 was furnished. But there was no
information regarding stratigraphy, so that until a final publication appears
with full details the incomplete information presented here remains tem-
porarily without significant [stratigraphic] meaning.
The aim of the following catalogue is to make available for study and
discussion that group of artifacts existing in one of the several reposito-
ries of the Surkh Dum corpus. And it will be clear that the catalogue does
not impinge upon the final report, it only anticipates it. The bronzes are
presented first, followed by the bone and ivory objects, the faience vessel,
and, finally, the cylinder seals, which are described and discussed by Eliz-
abeth Williams-Forte. The headings contain the Museum and Surkh Dum
record numbers. For each object a description is given along with a brief
48 The Museums records do not reveal in what capacity Pope functioned when he sold
the objects.
402 chapter twelve
commentary on function, and, where available, parallels among both exca-

vated and unexcavated material. The latter group is of particular impor-
tance, for the Surkh Dum example now anchors them (or some) objectively
within Luristan. Concerning the objects excavated elsewhere in Luristan,
not only is a Luristan provenience reinforced, but multiple site proveniences
are demonstrated; and if the parallel piece was excavated outside of Luris-
tan, cultural relations between two distinct areas are documented, again
objectively (e.g., No. 31 and the cylinder seals). In some instances I refer to
the existence of unpublished objects from Surkh Dum. This information is
derived either from notes in the possession of Trudy S. Kawami, who gener-
ously shared them with me, or from observations made by me in the Oriental
Institute, Chicago, and in the University Museum, Philadelphia.
Catalogue
Metal Artifacts
No. 1. Plaque. 43.102.11; Surkh Dum 1721. Bronze; P.H. 9.5 cm.
Two stylized leonine creatures face each other in heraldic position. Their
mouths are open and tongues and fangs protrude; manes are depicted as
thick curling tufts. The upward curling tail, the neck and the thigh on the left
figure are accentuated with punched dots. The paws apparently originally
touched and both creatures seem to be rampant, standing on rear legs. The
border consists of raised dots framed by narrow bands; the top and part of
the right border are extant.
There is at least one other fragment of a similar plaque from Surkh Dum,
preserving only the eye and mane of a creature exactly like the present
example. I know of only one parallel to these two plaques, one more com-
plete, and formerly in the David-Weill collection.49 From this example we
are able to restore our fragments as a rectangle, with perhaps a dead ani-
mal under the creatures, indicating perhaps that the latter are fighting over
their prey. The David-Weill plaque has subsidiary motifs around the crea-
tures lacking on ours; what is more, those lions are winged and the necks
and bodies are rendered in a more baroque fashion than ours, indicating a
separate workshop. What function the plaques had is not known, but, inas-
much as there are no holes along the borders, we may presume that they
were not meant to be attached to leather or another backing. The creatures
are not typical Luristan types and thus add another dimension to the reper-
tory.
No. 2. Plaque. 43.102.12; Surkh Dum 1269; Plot JH R2. Bronze; L. 5 cm.; H. 2.5 cm.
This small, thin rectangular plaque has no holes for attachment to another
object or material and is complete as is. A recumbent, horned animal in low
relief and with no body decoration faces right. Its legs are tucked under its
body, and its hooves touch. The animal is framed by raised dots.
Whatever function plaques like this and No. 1 had at Surkh Dum is as yet
unknown, pending the publication of their original find spots. Recumbent
49 Pope 1941, op. cit. (in note 7) 293, fig. 7; Amiet 1976, op. cit. (in note 1) no. 196: which may
now be attributed to Luristan with some security.

404 chapter twelve
animals with their feet and hooves in the same position are to be seen on
disc-headed pins and other Iranian objects attributed to Luristan, as well as
from Marlik and on objects from Elarn;50 the image is thus well established
within Iran. A plaque published by Godard51 that depicts recumbent birds
may have had a similar use and meaning: votive plaque?
No. 3. Disc-headed pin. 43.102.10; Surkh Dum 858. Bronze; D. 5 cm.

Extant is a human face in repouss high relief, probably that of a female,
with only fragments of the surrounding background disc; the pin, originally
hammered from the same sheet of metal, is missing. The face is round,
the mouth, thin and lunate shaped, appears to be smiling; herring-bone
decorated brows meet over the flat, broad nose; eyes are almost almond-
shaped and have no pupils; the hair, parted at the middle, consists of incised
ovals with a punched dot.
50 Luristan: Amandry, op. cit. (in note 34) plate XXVI, 3; Clercq-Fobe, op. cit. (in note 21)
nos. 27, 40, 41. Marlik; Ezat Negahban, Preliminary Report on Marlik (Teheran 1964) pl. VII A;
Elam: Amandry, ibid., pl. XVII, XXVIII; Pierre Amiet, Appliques iraniennes, Revue du Louvre
(1977) 6465, fig. 4.
51 Godard 1962, op. cit. (in note 21) fig. 35.
In Schmidts report no specific information was given about the disc-

headed pins except the statement that some pins were found sticking in the
walls of the temple. From the records at the Oriental Institute, however, it is
known that a large number of disc-headed pins with a variety of decorative
scenes depicting floral, animal, and human-like figures were excavated at
Surkh Dum. Van Loon claimed in 1967 that those pins with human faces
occur in the 8th century level, a date revised in 1972 to the late 8thearly 7th
centuries bc. Two sub-groups of these particular pin types occur at Surkh
Dum (and among the many stray finds), one where the face occupies almost
the whole disc, as the present example, and those where the face is placed at
the center and is encircled by either geometric or floral motifs, or by animals
or human-like figures.
The human faces have been assumed by some scholars to represent
a deity, while a few scholars see some as deities and others as human
portraits.52 The former opinion seems to me to be more in context with
what we would expect in ancient Near Eastern art. There is also confusion
52 Deity: Dussaud, op. cit. (in note 23) 200; Ghirshman 1956, op. cit. (in note 21) 196; Moorey
1971, op. cit. (in note 6) 214215. Deities and humans: Godard 1962, op. cit. (in note 21) 6364;
Clercq-Fobe, op. cit. (in note 21) 22, 40.
406 chapter twelve
concerning whether all the heads are those of females, or some are females,
others males.
Aside from the pins from Surkh Dum, a large number with a variety
of motifs, many formally matched among the excavated examples, have
surfaced on the antiquity market since the first time Luristan bronzes began
to appear about 1930 (see note 31). Parallels for the pin under discussion,
where the human face occupies the whole disc, may be found in addition to
those from Surkh Dum itself among several stray examples.53 On the basis of
the little information published by Van Loon about the site of Surkh Dum, it
appears that the disc-headed pins do not pre-date the late 8th century and
may actually continue into the 7th century bc.
No. 4. Pendant. 43.102.5; Surkh Dum 419. Bronze; D. 8.8 cm.

This thin, fragmented bronze sheet metal has a rolled loop at the top indi-
cating that it is a pendant, not a pin. In repouss is depicted a human-like
male figure kneeling on one knee, in the knielauf position, right. His head,
en face, is bearded, but no mouth is shown; ears are large and pointy; eyes
are simple bulges, and the nose is broad and flat. The body seems to be nude
although no sex is depicted. In each hand are held palm fronds (?).
I do not know if there are other pendants or disc-headed pins with
a similar scene in the Surkh Dum repertory, but a number of stray disc-
headed pins depict basically the same figure and motif. On each example
there is depicted a hybrid human-like figure, apparently always male, in the
knielauf position to the left or right, en face, sometimes horned as a bull or
caprid, and sometimes apparently clothed, and always holding objects in
each hand. On at least one example the figure holds snakes; on one pin he
holds pomegranates; on another, where the penis is a pomegranate, he holds
an animal and a bird; on two pins he masters two animals; and on another
example he holds unidentifiable objects.54 A problem exists, of course, with
53 Ghirshman 1956, op. cit. (in note 21) pl. XXIV; Godard 1962, op. cit. (in note 21) figs. 6871,
73, 74; Clercq-Fobe, op. cit. (in note 21) nos. 2932.
54 A.U. Pope, Prehistoric Bronzes of a Hitherto Unknown Type , ILN (May 6, 1939) 790,
fig. 7; Godard 1962, op. cit. (in note 21) figs. 34, 36; Clercq-Fobe, op. cit. (in note 21) no. 50; Peirre
Amiet, Les Bronzes du Luristan, Revue du Louvre (1963) 16, fig. 8; idem, Notes dArchologie
iranienne, Revue du Louvre (1969) 328, fig. 5. Because of the extraordinary variety of forms and
motives, and different workshops involved, it is always a difficult and complex problem to
separate the genuine from the forgeries among the many unexcavated works of art attributed
to Iran. Thus, in Muscarella 1977, op. cit. (in note 37) 173, no. 25 I listed as suspicious a disc-
headed pin in Geneva depicting a kneeling demon holding snakes. After restudying this pin
in context with others depicting kneeling figures I must retain my doubts: note that the
regard to whether one particular deity or genius, or several, were repre-

sented, especially given the variety of the objects, no doubt attributes, car-
ried, and the presence or lack of horns. The usual interpretation is that
the figure represents a deity or genius connected with fertility, especially
because of the presence, in one case significantly, of pomegranates, and
snakes.55
Geneva disc is thick and may not be a pin head (see Bernard Goldman, The Asiatic Ancestry
of the Greek Gorgon, Berytus XIV [1961) 19, plate I, 1, who thinks it is a mirror). Note the
sharp outlines of the demons body and the sharp demarcation line of the projecting right
leg separating it from the body, and note the drawing of the feet, features not present on the
Surkh Dum figure, nor on other examples. The head of the demon also bothers me, especially
when seen from the rear. I also do not cite a disc-headed pin published in Mostra dArte
Iranica (Rome 1956) plate XVI, left (Muscarella 1977, ibid., no. 20). Here I am not convinced by
the drawing of the demons hair, eyes, face in general, and the presence of the pubic triangle,
as well as that of the two animals; it is at least suspicious.
55 Clercq-Fobe, op. cit. (in note 21) 31, 120. Goldman, op. cit. (in note 54) 1, 3, 5, 69, sees
the Geneva demon mentioned in note 54 to be a bearded female, an adrogynous creature,

408 chapter twelve
The David-Weill collection56 at one time contained a small bronze group

in the round that may have once surmounted a pin: a central figure, almost
in a knielauf position, en face, holds at bay two lions. The figures posture,
face, and apparent lack of clothing are reminiscent of the figure on our
pendant and on the pins and may represent the same personage.57
Moorey suggested that the origin of the disc-headed pins may have re-
sulted from an adaptation of circular pendants. While this cannot be proven,
the existence of this pendant with a motif matched on the pins does not
contradict his hypothesis. Note also a possible disc pendant in Brussels.58
No. 5. Open-work cast pin. 43.102.1; Surkh Dum 1573; Plot J1 pr 178/45.59 Bronze;
L. 14.3cm.
Cast in one piece, this open-work pin depicts a squatting female en face,
her legs spread and touching the frame. Small pellet breasts and exposed
pudenda identify the figure as a female. The face and head are corroded but
one may see facial features: small eyes, lips, as well as the horns that identify
her as a deity. Earrings or spiral hair locks exist on either side of the face, and
a grooved area above the face may represent hair. The female is nude but
there is a grooved rectangular area above the pudenda that may represent
a girdle. Held at bay by her thin, unnaturally curved arms are two horned
animalsantelopes/goatsstylistically rendered only by their heads and
long, thin necks that join in a continuous curve, and which enclose the
deity in a frame; unidentifiable curved units connect the animal heads to
the deity. A non-descript thin unit joins the frame to the pudenda, but it is
not clear whether it is a strut or had a more significant meaning, namely
representing the process of birth.
Open, cast pins depicting either a mistress or master of animals are one of
the most characteristic forms among the Luristan bronzes. The iconography
occurs on many pins but is not limited to them, for, among other items, it is
who is a Great Mother-Lamashtu-Gorgon figure, an interpretation also held for the other
kneeling figures on the pins. There is indeed a formal parallel between the Geneva demon
and the Larnashtu depictions, but if the former is not ancient, the idea collapses. Moreover,
the conclusions regarding the sex and attributions of the kneeling figure are to my mind not
so secure as presented.
56 Amiet 1976, op. cit. (in note 1) no. 186.
57 Cf. a pin formerly in the Bach collection, Bronzes de la Perse (Paris, Hotel Drouot
12/12/1973) no. 28, and Godard 1962, op. cit. (in note 21) fig. 81.
58 Moorey 1971, op. cit. (in note 6) 208; Clercq-Fobe, op. cit. (in note 21) 223, no. 26.
59 Listed in Ackerman 1940, op. cit. (in note 4) 543, G. She called the unit joining the
pudenda to the frame an exaggerated Phallus.

of course characteristic of the classic Luristan finials. The specific iconogra-

phy of the Surkh Dum pin is primarily the squatting position of the female
mistress of animals with her sex exposed, and a basic shape where the necks
of creatures curve so that the deity is enclosed within a crescentic frame.
The type has been discussed by Moorey,60 and many stray examples exist:
our Surkh Dum pin is to date the only one excavated (it is possible that this
pin is the very one faintly visible in situ in Schmidts report [p. 213, fig. 9]; I do
not know how many, if any, other examples of the type were found at Surkh
Dum).
Each pin of the type under discussion was cast by the lost-wax process
representing an individual modelling, and variations exist from one exam-
ple to the next. Thus, while some females have horns and spiral hair curls as
well as the grooved girdle, others may not; the animals, rendered in the typ-
ical Luristan manner, may be antelopes or lions.61 Sometimes the female is
60 Moorey 1971, op. cit. (in note 6) 204205.

61 Ghirshrnan 1964, op. cit. (in note 21) fig. 54, an elaborate example; Legrain, op. cit. (in
410 chapter twelve
represented standing, but the horns and spiral hair curls, and sometimes the
presence of the grooved girdle, identify her as the same or a related deity, or
essence, as the squatting female;62 and sometimes the female has no legs rep-
resented, or merely a head.63 The variety is large and pins of the type under
discussion are but one type of a large group reported from Luristan that is
clearly related iconographically: defined by the primary motif, a mistress or
master of animals framed either in a crescent, or within a square or circular
unit.
The squatting female motif is not limited to open-work pins, for at least
two examples are known depicted in relief on disc-headed pins, in one
case where the female is actually giving birth.64 The squatting position in
general, a birthing position for women, as well as the specific birth scene,
surely supports the interpretation that these pins, open work or in relief,
had a votive value associated with fertility rather than female sexuality per
se. And probably the standing females, including those with animal heads,
likewise were involved with fertility matters. As for the standing males,
linked with the females iconographically, they, no doubt, had a different
charged function.
Moorey65 has suggested that the open-work pins, some of which are large
and heavy with frames, may have served as icons rather than as garment
pins. He has rightly noted their stylistic and iconographical relationship
with the finials, which were probably icons. Perhaps he is right in essentials,
especially given their presence in the sanctuary of Surkh Dum, but it need
not follow that they were not also worn on the body as a charm or protective
amulet, while also functioning as clothing fasteners.
note 31) pl. V, 14; Moorey 1971, op. cit. (in note 6) no. 348, a variant with a square frame; Amiet
1976, op. cit. (in note 1) nos. 178, 180, the former with a mouflons head over the female, and
exactly matched by a pin in the Khoramabad Museum, Mina Sadegh-Behnam, Anita Koh,
Lorestan Bronzes and Islamic Metalwork (N.D.) no. 4.
62 L. Speleers, Nos Nouveaux Bronzes perses, Bulletin des Muses Royaux dArt et dHis-
toire (1932) 102, fig. 27. Other bronzes in the Brussels Museums collections are published in
the same journal and under the same title for the years 1932 and 1933: (1932) 5671, 93104,
115119; (1933) 6269, 8695. Cf. similar pieces in Moorey 1971, op. cit. (in note 6) no. 346, and
J.A.H. Potratz, Luristanbronzen (Istanbul 1968) no. 101.
63 Amiet 1976, op. cit. (in note 1) no. 178; Moorey 1971, op. cit. (in note 6) no. 347; Godard
1931, op. cit. (in note 31) pl. XXXV, 150; H. Potratz, Das Kampfmotif in der Luristankunst,
Orientalia 21:1 (1952) 2628, figs. 2138, 41, illustrates a number of these pins of different types
and considers them to represent a Luristan moon goddess.
64 Amiet 1976, op. cit. (in note 1) no. 189; Godard 1962, op. cit. (in note 21) figs. 77, 78: the
former pin was once in the Godard collection, not David-Weill as thought by Moorey 1971, op.
cit. (in note 6) 204.
65 Moorey 1971, op. cit. (in note 6) 200; idem, 1974, op. cit. (in note 6) 124.
No. 6. Pin. 43.102.7; Surkh Dum 1539; Plot JI pr 178/11.66 Bronze, iron; P.H.
3.8cm.; W. 7cm.
A central motif is flanked by the heads and long necks of two stylized
Luristan type felines. The necks are joined and continuous and half-way
enclose the central unit in a frame, similar in form to some open-work pins
(see No. 5). Seen from the front, it is difficult to recognize what the central
motif represents; seen from the side, it is easily recognized as a ducks head
turned back to recline on its wings (cf. the bracelet from Bard-i-Bal).67 The
shank of the pin, now missing, was made separately of iron, judging from
the color at the join. A number of examples of this pin type, some framed
with feline, others with antelope, heads, occur at Surkh Dum.
I know of only two published parallels to our pin.68 Surkh Dum also
yielded many examples of straight pins with the very same central motif as
on ours but lacking the flanking heads and necks; other examples exist in
various collections.69 These pins are now confirmed as being from Luristan,
and the occurrence of iron surely attests to a date not earlier than the late
9th century bc, and probably later.
66 Listed in Ackerman 1940, op. cit. (in note 4) 544, P, and incorrectly described.
67 Louis vanden Berghe, La Necropole de Bard-i Bal, Archeologia 43 (1971) 21, fig. 15.
68 Speleers 1932, op. cit. (in note 62) 102, fig. 28, and Hotel Drouot Catalogue, 5/22/80,
no. 273 bis; cf. L. Speleers, Antiquits iraniennes, Bulletin des Muses Royaux dArt et dHis-
toire (1938) 42, fig. 16: is this a pastiche?
69 Godard 1931, op. cit. (in note 31) pls. XXXIII, LVI, 119, 125, 205; F. Basmachi, The Luristan
Bronze Objects in the Iraq Museum, Sumer XIX (1963) pl. 6; Moorey 1971, op. cit. (in note 6)
no. 317.
412 chapter twelve
No. 7. Amthropomorphic pin. 43.102.6; Surkh Dum 1207; Plot JI R3 pr 163.70

Copper; H. 6.4cm.
The solid cast head of this pin is rendered in a stylized manner, suggesting
as an impression that a large head alone is depicted, when in fact a whole
figure is sculpted. The head clearly is meant to predominate and apparently
represents a dominant figure. This interpretation is suggested by the promi-
nent, sharp nose, thick brows encircling the eyes, and either a large lantern
jaw or a beard clearly offset from the mouth area; no ears are depicted. A
sloping, flat beret-like cap does not completely cover the hair, for a band of
wavy locks or curls runs across the forehead. The head joins a tubular sec-
tion that functions as both neck and body, albeit that arms are not depicted.
At the base of the body are thighs and legs in a squatting position, and
the feet grasp the top of the shank on either side. The impression is that the
figure sits on a pinnacle, holding on by the feet. Breasts are not indicated,
but between the knees is a raised oval area with a central depression, which
70 Listed in Ackerman 1940, op. cit. (in note 4) 546, NN.

suggests that it is a vulva. If, however, the figure has a beard and not a long
jaw, then we have something else here. Note that laboratory analysis has
determined that the figure is made of copper, not bronze.
I can find no parallels for this figure, for its position, or face and hat.
It is further unique in that it is at present the only published example of
a figure in the round to have been excavated in Luristan. The literature is
filled with examples of stray bronze (copper?) figurines claimed to derive
from Luristan, but they are all standing figures and none has the armless
neck-body arrangement as ours. This latter feature, however, does seem to
exist on the demon figure on the many finials reported from Luristan, and
in one small but significant detail they present another parallel to our piece:
on a few finial examples it seemsone is not certainthat the neck-body
has feet that grip the base below the neck section.71 If this observation is
correct then we have another formal connection between our figurine and
the typical Luristan finial demons.
No. 8. Human-headed pin. 43.102.17; Surkh Dum 44; Plot J1 N.E. Bronze; H.
3.5cm.
71 Viz. Potratz 1968, op. cit. (in note 62) pl. XXXVI, XXXVII, nos. 228231 (note that on
nos. 232, 234, 235, 238, 242, the visible feet may belong to the heraldic creatures).
414 chapter twelve
That this is a pin is indicated both by the presence of a shank, now mostly
missing, and because there is another, more complete example from Surkh
Dum (SOR 201) now in the University Museum, Philadelphia. Our example
is topped by a small, beardless human head with a prominent nose and thick
lips, centered on a curved unit that projects on either side and which may
represent either arms, otherwise not represented, or more probably, wings.
On the Philadelphia example the wings are longer and thinner, and curve
up in a pronounced manner at the tips. At Surkh Dum there are examples
of pins exactly like ours in form except that instead of the head at the center
there is a short, raised scalloped unit.
Outside of Surkh Dum no other examples of this pin type with a human
head are known. On the other hand, an example with the raised central
scalloped unit, exists in Brussels.72 And clearly related examples, with a plain
small central unit, exist in the Ashmolean Museum and among the bronzes
claimedwithout verificationfor Khurvin;73 another related type, one
with a swelling at the central part of the wings, is also claimed for Khurvin.74
A more developed type, perhaps also related to our pin in concept, depicts
the torso of a male centered on the wings, or a male torso centered on rams
horns.75
If the projections on our pin, and the others, are indeed wings, one would
be right to assume that we have a representation of a deity, unnamed to be
sure, but appropriate for dedication at a sanctuary.
No. 9. Animal- and demon-headed pin. 43.102.3; Surkh Dum 209; Plot JI pr
2597.90.76 Bronze; P.H. 3.5cm.
The head of this pin consists of a stylized head of a demon or deity sur-
mounted by a recumbent animal; it is meant to be viewed from the front
only. The face is formed by pellet eyes encircled by a thick line that is the
brows and that continues to create the nose; no ears or mouth are depicted.
Two loops placed on top of the head may represent hair curls and curved
72 Speleers 1933, op. cit. (in note 62) 89, fig. 26.
73 Moorey 1971, op. cit. (in note 6) no. 289; Louis vanden Berghe, La Ncropole de Khurvin
(Istanbul 1964) pl. XLIII, no. 316. Note that many of the objects published as from Khurvin
derive from a Teheran private collection.
74 VandenBerghe, ibid., no. 314; see also Godard 1931, op. cit. (in note 31) pl. XXXIII,
129; cf. A.U. Pope, ed., A Survey of Persian Art (Tokyo 19641965) pl. 59 D; Wolfram Nagel,
Altorientalisches Kunsthandwerk (Berlin 1963) pl. VIII, 18.
75 Ernst Herzfeld, Iran in the Ancient East (London and New York 1941) 155, fig. 275, center
right; Moorey 1971, op. cit. (in note 6) no. 342.

76 Listed in Ackerman 1940, op. cit. (in note 4) 545, X.
units encircling the face may be horns. An animal, whose head is now
missing, rests on the head of the demon, its feet touching the sides. Under
the grooved moulding is a hole that once held a separately made pin; it is no
longer possible to know if it was bronze or iron.
To my knowledge, no parallels for this specific pin exist, but there are
at least two examples that in form closely relate to ours. These pins have
what Porada has called two profile lion heads (which also combine to form
a single frontal mask).77 On both these pins the masks, if that is what they
77 Porada 1979, op. cit. (in note 44) 142143, note 17, fig. 9; A.U. Pope, Mute, Yet Eloquent:
The Significant Luristan Bronzes , ILN (September 13, 1930) color plate, lower left.
416 chapter twelve
actually are, are surmounted by a mouflon with large, majestic horns. While
the legs on the Surkh Dum example are placed on either side of the central
mask, however, on the other pins they rest directly on top. It may be that our
animal also originally had a head with the same majestic horns.
An exact parallel for the stylized mask with top loops on our pin exists
in relief on a disc pin in Brussels, and in the round on a handle (?) in the
Erlenmeyer collection.78
No. 10. Animal-headed pin. 43.102.8; Surkh Dum 1078.79 Bronze; P.H. 5.7 cm.
Cast in the round, the head is in the form of a standing goat whose feet are
drawn together to rest on a plinth. Below the plinth is a tubular moulding
into which a separately made pin, of unknown material, was originally
inserted.
78 Clercq-Fobe, op. cit. (in note 21) 208, no. 19, pl. 19; M.L. and H. Erlenmeyer, Frhi-
ranische Stempelsiegel, II, Iranica Antiqua 5 (1965) 8, pl. V, 24, pl. XI; cf. similar heads in
J.A.H. Potratz, Die Luristanbronzen des Museums fr Kunst und Gewerbe in Hamburg,
ZAssyr 17 (1955) pl. I, 4.
79 Listed in Ackerman 1940, op. cit. (in note 4) 545, W.
Goats and mouflons in the round were very commonly depicted on the
art of Luristan; on finials, horse harnesses, pins, and so forth. With regard to
goats on pins, there is a group that often represents the animal either stand-
ing or recumbent with the head turned toward the viewer, as on horse cheek-
pieces;80 our example is distinct in that the goat faces forward. Characteristic
of most examples is the position of the feet, which are drawn together as if
balanced on a point, a mountain peak.81
Pins of this type may have been icons rather than, or in addition to, being
used as clothing fasteners and charms.
No. 11, Frog-headed pin. 43.102.4; Surkh Dum 400; Plot II pr 24.82 Bronze; L.
4.8cm.; W. 2.5cm.
The strange creature seems to be a frog depicted in the round and as seen
from the top. Its eyes bulge, and all four legs project in the same direction;
the body is simply rendered except for the back ridge that connects the legs.

81 Cf. Godard 1931, op. cit. (in note 31) pl. XXXV, 147; Legrain, op. cit. (in note 31) pl. V,
17; Pope 19641965, op. cit. (in note 74) pl. 60 F; Peter Calmeyer, Altiranische Bronzen der
Sammlung Brckelschen (Berlin 1964) no. 127, cf. no. 128 for posture. Cf. also goats on plinths
discussed in Oscar White Muscarella, The Archaeological Evidence for Relations Between
Greece and Iran in the First Millennium B.C., |it Journal of the Ancient Near Eastern Society
9 (1977) 34, note 11, fig. 7; Amiet 1976, op. cit. (in note 1) no. 198.
82 Listed in Ackerman 1940, op. cit. (in note 4) 546, 00. The pin is published by Paul
Jacobsthal, Greek Pins (Oxford 1956) 61, no. 257.

418 chapter twelve
The frog is clearly the head of a pin, the shank of which is cast with it and
part of which is still extant.
A similar pin, complete, ending in a frogs body was at one time in the
David-Weill collection;83 it was made by a different hand than the one that
made ours and has a loop at the shanks base to hold a chain or cord to
facilitate securing the pin to a garment. In addition, the Boston Museum
of Fine Arts has an amulet in the form of a frog that was purchased in 1930;84
a frog amulet also is in Baghdad.85 Although apparently rare, the frog is thus
attested as a decorative element for pins and amulets in Luristan. Aside from
our Surkh Dum example one or two more were found at Surkh Dum; no
others, either pins or amulets, have been excavated.
No. 12. Animal-headed pin. 43.102.20; Surkh Dum 1432; Plot J1 R3 pr 175/63.
Bronze; L. 6.6cm.
83 Amiet 1976, op. cit. (in note 1) 73, no. 173.

84 Pope 19641965, op. cit. (in note 74) pl. 59 C; Muscarella 1979, op. cit. (in note 1) 12,
note 10: see Pope 1930, op. cit. (in note 77) 444, fig. 4.
85 Basmachi, op. cit. (in note 69), pl. 11, bottom right.
The head is formed with two individual units: a rounded band, decorated
in low relief with studs or cast granulations framed by beaded mouldings,
which is joined to the head of an animal, whose head, horns, and ears
are rendered naturally in the same plane as the shank. Under the head is
a groove running its whole length and with its side arms grooved at the
nose end. This feature, seen in side view, could indicate the animals legs
and the swelling behind the eyes could be its shoulders. Thus we have the
forepart of an animal, not just its head; this feature distinguishes this pin
from the antelope-headed pins, Nos. 1417. From a formal point of view,
one thinks of course of the zoomorphic straight-headed pin from Baba
Jan,86 where the whole body of a typical Luristan feline creature forms the
head.
No. 13. Duck-headed pin. 43.102.19; Surkh Dum 423; Plot II pr 2425 98.00.87
Bronze; L. 20.6cm.
Cast in one piece, the head is in the form of a reclining duck that is separated
from the shank section by a series of grooves. In addition to this pin and
others from Surkh Dum, the only other site that has yielded this type pin is
Kutal-i-Gulgal, also in Luristan.88 A large number of stray examples exist, all
exactly the same as those from Surkh Dum and Kutal-i-Gulgal,89 and which
now have a confirmed Luristan provenience.
The presence of the same type of object both at Surkh Dum and at an
excavated cemetery site in Iran is archaeologically significant. First of all,
aside from gaining knowledge about distribution, it demonstrates that, as
in the present case with regard to pins, the object could have a votive and
86 Clare Goff Meade. Luristan in the First Half of the First Millennium B.C., Iran VI (1968)
128129, fig. 12.

87 Listed in Ackerman 1940, op. cit. (in note 4) 537, HH.
88 Vanden Berghe 1973, op. cit. (in note 44: Archeologia 65) 19, 21, 24.
89 Godard 1931, op. cit. (in note 31) pl. XXXIII, 137; Nagel, op. cit. (in note 74) pl. LVI, 124;
Basmachi, op. cit. (in note 69) pl. 5; Pope 19641965, op. cit. (in note 74) pl. 60 K; Potratz
1968, op. cit. (in note 62) 36, note 4, pl. XXIV, 140; Moorey 1971, op. cit. (in note 6) 193194,
nos. 314315; Anton Moortgat, Bronzegert aus Luristan (Berlin 1932) pl. VII, 19; W.D. van
Wijngaarden, De Loeristanbronzen in het Rijksmuseum van Oudheden, Oudheidkundige
Mededelingen XXXV (1954) pl. XlI, 75, 76.
420 chapter twelve
perhaps a secular function. And second, it serves as a warning that all stray
objects of a type usually related to Surkh Dum may not in fact have derived
from there; some of these pins cited above have been known since 1930.
The date of the Kutal-i-Gulgal tomb containing the pins was originally
dated by Vanden Berghe to a time around 11001000bc, but later the date
was modified and lowered by a century, to ca. 1000900bc: this is still earlier
by a century or more than the time it is suggested the Surkh Dum sanctuary
flourished (supra), and it creates a paradox. Either the Kultal-i-Gulgal tomb
is even later than suggested (see also No. 19), or the Surkh Dum sanctuary
begins earlier than perceived, or, a third possibility, the pins at the latter site
are heirlooms or long lived. In any event, all that can be stated at present
is that there seems to be a considerable difference in time between the
occurrence of the pins at two excavated sites.
Four other animal-headed pins (Nos. 1417) are exactly the same in all
details, differing only in size and horn positions (indicating that they were
made by the lost-wax process). All terminate with the head of a horned
animalan antelope? The horns are free from the head and pass between
the upright ears. The pin shanks vary in length and thickness; No. 17 is bent
and No. 16 is broken.
A very large number of antelope-headed pins were excavated at Surkh
Dum and an equally large number of strays have been recorded from the
time of the earliest appearance of Luristan bronzes on the antiquities mar-
ket.90 The former examples thus neatly confirm a Luristan provenience for
the class. Note that No. 17 is specifically mentioned as having been found in
a wall.
No. 14. 43.102,18; Surkh Dum 1203. Bronze; L. 14.5 cm.
90 Moorey 1971, op. cit. (in note 6) 193 for references, nos. 312, 313; see also Godard 1931,
op. cit. (in note 31) pl. XXXIII, 123, 132, 133; Herzfeld 1941, op. cit. (in note 75) 155, fig. 275;
Wijngaarden, op. cit. (in note 89) pl. XII, 7780; Calmeyer 1964, op. cit. (in note 81) no. 67;
Jean Paul Barbier, Bronzes iraniens (Geneva 1970) no. 41.
No. 15. 43.102.21; Surkh Dum 279; Plot J1 pr 1997.45. Bronze; L. 13.9 cm.
No. 16. 43.102.22; Surkh Dum 197; Plot II. Bronze; L. 8.3 cm.
No. 17. 43.102.23; Surkh Dum 1585 or 1685; Plot KH, in the wall. Bronze; L.
9.2cm.
No. 18. Pin. 43.102.24; Surkh Dum 578; Plot KI pr 11.91 Bronze; L. 12 cm.
The head consists of a double row of small projecting knobs, apparently

meant to form rosettes, set on a grooved base. Laboratory tests have indi-
cated that the pin is made of tin bronze.
Van Loon mentions pins at Surkh Dum from the earliest level; he referred
to studded pin heads in the 8th7th century level,92 which may describe
the pin type under discussion here. Although the publication of pins is
extensive, reflecting both their variety and ubiquity, I could find no exact
parallels to our example, but compare the rosette-like motif on a silver pin
claimed for Ziwiye.93
91 Listed in Ackerman 1940, op. cit. (in note 4) 543, I.

93 C.K. Wilkinson, More Details on Ziwiye, Iraq XXII (1960) pl. XXX, 5.
422 chapter twelve
No. 19. Animal-terminal bracelet. 43.102.2; Surkh Dum 1632. Bronze; D. 8.2 cm.
The terminals on this cast penannular bracelet are in the form of the forepart
of a stylized animal of indistinct species (lion?); on each back is a loop,
probably used to hold a cord or chain to keep the bracelet from slipping
off the wrist; the arms are round in section and plain.
One of the most characteristic and numerous of the objects reported
from the plundered tombs of Luristan is the bracelet with zoomorphic
terminals, hundreds of which have surfaced through the antiquities market.
Bracelets of this type are rare from excavations in Luristan, but it is known
now that they exist not only at Surkh Dum (i.e., the present example and
others reported), but also at Bard-i-Bal,94 so the Luristan provenience for
the class in general is documented. Moorey95 has discussed the importance
of these objects, both with respect to their earlier occurrence in Iran at
Hasanlu and Marlik, and equally significant, their flourishing continuity
in Achaemenian times. Concerning the latter occurrence, animal-terminal
bracelets represent one of the most clearly documented examples of earlier
Iranian art forms taken over and developed by the Achaemenians.
94 Vanden Berghe 1971, op. cit. (in note 67) 21, fig. 15; idem 1973, op. cit. (in note 44: Iranica
Antiqua) pl. XXI, 1, 2.

Aside from the Surkh Dum examples, no other bracelets like our example
with terminals in the form of the whole or the forepart of an animal (as
opposed to the many with only the animal head) have been excavated, but
strays exist.96
No. 20. Lobe-shaped ring. 43.102.13; Surkh Dum 1601.97 Bronze; D. 2.5 cm.
No. 21, Lobe-shaped ring. 43.102.14; Surkh Dum 10298 Bronze; D. 2.3 cm.
Ring No. 20 was made from a sheet of bronze wider in front than at the
back where the slightly narrower ends touch. The design, which may have
been cast with the sheet, consists of two different horned animals, one
with a curved horn shown in profile, the other with two horns shown
frontally, flanking a stylized tree; a multi-petalled rosette or star is behind
each animal. In the field is an inscription in cuneiform which reads Dinger.
Mesh, part of a prayer invoking the gods. The second ring is similar in shape
96 Godard 1931, op. cit. (in note 31) pl. XXVIII, 94; Pope 19641965, op. cit. (in note 74)
pl. 57 C; Ghirshman 1964, op. cit. (in note 21) fig. 94; Louis vanden Berghe, et al., Bronzes
Iran-Luristan Caucasus (Paris 1973) pl. XXXVI, center.
97 Listed in Ackerman 1940, op. cit. (in note 4) 547, QQ.
98 Listed in ibid. 547, RR.
424 chapter twelve
to the first, although cast closed and with a more pronounced lobed front.
The design, while neat, is executed in eneven lines and seems to have been
added after casting. A bird facing right is placed above a horned animal
striding left with its head down: on the sealing the positions are reversed.
Along the outside borders is an incised line. Both rings may have been used
as seals.
Edith Porada99 has made a study of these rings, seeing them as stylistic
indicators for establishing a chronology for certain Luristan bronzes. She has
categorized rings of our first type as sheet rings, those of the second type as
lobed rings. She has also perceived a chronological distinction between the
two types, exhibited by the fact that the engravings on the sheet examples
are usually carefully rendered and often show heraldic animals flanking a
stylized tree. To Porada, the scenes on the lobed rings are usually cruder in
execution and were added after casting. While not explicit, it seems that
she considers those rings cast with the arms closed to be lobed, and the
penannular ones to be the sheet rings.
Inasmuch, however, as some lobed rings have penannular ends and
finely executed scenes, and because the sheet rings are all lobed in shape,
the division is not so clear as assumed. It may be that the type and style of
the scenes themselves should be the criteria for the division among the lobe-
shaped seals: the neatly rendered heraldic animals and tree, all of which
seem to have penannular ends, and the others, some crude others neatly
rendered, with animals or demons not in a heraldic position and usually in
a different style than the first grouping.
The scenes on the sheet rings have been compared by Porada to Elamite
and Babylonian art and dated accordingly to 12001000bc; the lobed rings
are dated ca. 1000800bc.100 Van Loon, while noting that the relative se-
quence suggested by Porada, that sheet rings precede lobed rings, is sup-
ported by the Surkh Dum stratigraphy, has nevertheless claimed that the
former occur there in the 8th century level101 (= 8th7th centuries bc), the
latter in the 7th century bc level. Given the fact that early seals occur at
Surkh Dum, it is not impossible that Poradas date of the sheet rings to the
99 Porada 1964, op. cit. (in note 28) 1619; idem 1965, op. cit. (in note 28) 7578.
100 In Porada 1964, ibid. 16 the sheet rings are dated 1100 bc; later, 28, about the twelfth
or eleventh century bc; the lobed ones are dated on page 17; see 1965, ibid. 76, figs. 47, 48, 78.
Cf. Erlenmeyer, op. cit. (in note 78) 25 for a different arrangement and dating, but based on
gratuitous comparisons and to my mind not convincing.
late 2nd millennium bc is correct: yet one has to accept a 200400 year
difference between the two ring types, which in shape at least are not so
different. How many lobe-shaped rings were excavated at Surkh Dum is still
not revealed, but it would be of value to know if in fact sheet rings occur only
in one level, lobed rings only in another.
Thus, the issue of dating remains to my mind still unresolved, especially
with regard to the so-called sheet examples: we have a situation where
style suggests an early dating and stratigraphy a later one. Porada has also
noted that penannular rings of sheet metal and with lobed faces, some of
gold, occur in other areas of the Near East in the late 2nd millennium bc;
very distinctly lobed rings of iron and bronze have also been excavated at
Hasanlu of 9th century bc date.102
No. 22. Pendant. 43.102.15; Surkh Dum 617. Bronze; L. 2.9 cm.; H. 2.2 cm.
This small cast pendant seems to depict a dog. Its raised tail curves up and
forward above a flat rear end; eyes are small raised pellets and the ears are
small; a suspension loop connects the neck and back.
The number and variety of pendants at Surkh Dum (see also No. 23)
is still not known but few are known elsewhere from excavations. Aside
from a bird pendant from Bard-i-Bal (see No. 23), pendants in the form of
102 Porada 1964, op. cit. (in note 28), 16; for Hasanlu see Aurel Stein, Old Routes of Western
Iran (London 1940) 398, pl. XXV, 2; others, unpublished, are known from the recent exca-
vations. Cf. similar, but less pronounced, lobed rings from neighboring Dinkha Tepe, Oscar
White Muscarella, The Iron Age at Dinkha Tepe, Iran, MMJ 9 (1974) figs. 43, 52, nos. 133, 342,
620.
426 chapter twelve
animals, one perhaps that of a pair of dogs, were excavated at Hasanlu.103

Dog figurines were fairly common in Mesopotamia, where they seem to have
been associated in many instances with certain deities.104 Whether our dog
pendant was associated with a deity or had a simple secular function is not
known: except that it does come from a sanctuary.
While few pendants have been excavated, a large variety of stray exam-
ples are said to have derived from western Iran, especially Luristan. They
are in the form of humans, vessels, different types of birds and animals, and
so forth; dogs do not seem to be rare.105 The use of pendants at Surkh Dum
remains uncertain, for we do not know how they were worn, on the wrist,
neck, belt, or whether they could have functioned as simple decorative ele-
ments of secular jewelry, or only had a charged, apotropaic value.
No. 23. Pendant. 43.102.16; Surkh Dum 1013. Bronze; L. 0.6 cm.; H. 1.9 cm.
The pendant seems to depict a reclining bird, apparently a duck. It is very

simply rendered with no details articulated. The base is flat and is incised
with a cross-hatch design; a suspension loop connects the neck and back.
Only one other bird pendant has, to my knowledge, been excavated to date,
recovered outside of a tomb at Bard-i-Bal in Luristan.106 Other examples exist
in private collections.107
103 A. Hakemi and M. Rad, The Description and Results of the Scientific Excavations at
Hasanlu (in Persian) (Teheran 1950) fig. opposite p. 72.

104 B. Meissner, Apotropaische Hunde, Orientalische Literaturzeitung XXV, 5 (1922) 201
202; W. Heimpel, Hund, Reallexikon der Assyriologie IV (19221975) 494497; Betty Schloss-
man in Ladders to Heaven, Oscar White Muscarella, ed. (Toronto 1981) 114116; Daphne
Achilles. in ibid. 201. For recent discussions see I. Fuhr in B. Hrouda, Isin-Isan Bahryat I
(Munich 1977) 135145.
105 Speleers 1932, op. cit. (in note 62) 115, fig. 10; Pope 19641965, op. cit. (in note 74) pl. 59 J;
Godard 1931, op. cit. (in note 31) pl. XXX, L.

106 Vanden Berghe 1973, op. cit. (in note 44: Iranica Antiquo) 48, pl. XXIII, 3.
107 Godard 1931, op. cit. (in note 31) pl. XXX, D.C. A; Basmachi, op. cit. (in note 69) pl. 11,
center; Moorey 1971, op. cit. (in note 6) 231232, nos. 416418.
No. 24. Miniature axe. 43.102.9; Surkh Dum 1500.108 Bronze; L. 4.8 cm.
In miniature size, this axe duplicates full-sized examples known both from
excavations and the antiquities market. Characteristic of this particular type
is both the cut away, slanted lower part of the socket, and the flange-butt
with a horizontal ridge; the socket and the flange have a thick outline. Exam-
ples of this type have different blade shapes that define them as chisels,
picks, or axes, but the slanting socket and flange interrelate the group as
belonging to the same class or type. Our example is an axe, the upper edge
horizontal, the lower curving up to the socket.
The type was studied by Maxwell-Hyslop, Deshayes, and Calmeyer,109 all
of whom isolate the characteristics as well as give evidence for geographi-
cal distribution in the Near East. Full-sized examples have been excavated
at Til Barsip in north Syria, and in Iran at Susa and Kalleh Nisar in Luris-
tan.110 Other examples have been attributed to Nimrud and Tepe Giyan, but
without verification; and Calmeyer mentions two from Mari.111 Counting the
108 Listed in Ackerman 1940, op. cit. (in note 4) 547, XX.
109 K.R. Maxwell-Hyslop, Western Asiatic Shaft-Hole Axes, Iraq XI (1949) 99100, Type 9;
Jean Deshayes, Les Outils de Bronze de lIndus au Danube, I, II (Paris 1960) I, 166, II, 70, Type
A5c; Ca1meyer 1969, op. cit. (in note 1) 3234.
110 Calmeyer 1969, op. cit. (in note I) 33, fig. 32; Amiet 1976, op. cit. (in note 1) 9, fig. 5; Louis
vanden Berghe La Ncropole de Kalleh Nisar, Archeologia 32 (1970) 72.

111 Herzfeld 1941, op. cit. (in note 75) 126, fig. 243, c, pl. XXVII; Calmeyer 1969, op. cit. (in
note 1) 3435.
428 chapter twelve
present example, three axes of our type have thus been excavated in Iran.
All these examples are dated to the last centuries of the 3rd millennium bc.
There is no evidence at Surkh Dum that the site pre-dates the 1st millen-
nium bc, so that the presence of a 3rd millennium axe type is an anomaly.
One might assume that it is an heirloom, or a stray found by locals and dedi-
cated at the sanctuary, or an indication that some examples of the type were
still being made at a later date; we do not know which possibility obtained.
Miniature weapons, daggers and axes, while not rare in the ancient Near
East are not common either.112 The occurrence of a miniature axe at a sanc-
tuary indicates that it was dedicated as a model of a functioning weapon.
Note that another axe of a different type (see note 6) was also excavated at
Surkh Dum.
Ivory, Bone, and Faience Artifacts

No. 25. Plaque. 43.102.31; Surkh Dum 656 (or 666b?). Ivory; length: 9.5 cm.;
height: 3.8cm.
In fragments when found, parts of all four edges of this rectangular plaque
are preserved. On side A there is no defined upper border while the lower is
a band of vertical lines; on side B both upper and lower borders consist of
heavy four-petalled rosettes. On both sides the right and left borders have
an irregular but neat guilloche pattern, the centers and curves of which are
drawn with a compass. The upper edge has two drilled holes, the lower has
three.
112 Hans Bonnet, Die Waffen der Vlker des Alien Orients (Leipzig 1926) 71, fig. 9, from a
temple at Assur; Maxwell-Hyslop, op. cit. (in note 109) 119, 120; E.A. Speiser, Excavations at
Tepe Gawra (Philadelphia 1935) pl. XLIX, 3; Nagel, op. cit. (in note 74) nos. 41, 42, 100; Vanden
Berghe 1964, op. cit. (in note 73) pis. XLIV, XLV, 332, 335337; Pierre Amiet, Bactraine Proto-
historique, Syria LIV (1977) 106107, fig. 14; Calmeyer 1969, op. cit. (in note 1) 27, D is over
twice the size as ours and may not be a true miniature.
Side A is decorated with two identical bull men (the horns are clearly
part of their heads), en face and touching hands standing side by side; each
holds a creature at bay with his outside hand. The bull men have a triangular
face with a rectangular nose, slit mouth, thick round eyes that look compass
drawn with a central dot, and a beard that frames the whole face; vertical
lines between the horns may be hair and no ears are depicted. They are
dressed in a belted, calflength one-piece (?), short-sleeved jumper that is
fringed throughout its length and at the lower border; a vertical band runs
from the neck to the belt. Feet are visible below the skirt and seem to point
to right and left. The creature on the viewers right is an upright lion whose
body is outlined in bordered lines, as are the bull mens dress. The creature
to the left is larger, but because of a break is not readily identified. He seems
to rest on his haunch and his feet and paws, one with claws pointed up, the
other down, touch the bull men; this creature may be a bear, not a lion. The
decoration, except for the guilloche and eyes of the bull men, is rendered in
a crudely incised manner, especially the feet and hands of the bull men and
the bodies of the creatures.
Side B is equally rendered in a crude fashion. Here a lion at the left attacks
a horned animal fleeing to the right. One paw of the lion touches its prey,
both of whose front feet are off the ground, perhaps to show that it is falling.
The body outlines are like those on side A.
430 chapter twelve
The scenes on both sides are complete and the plaque is an individual
unit. Inasmuch as both sides are decorated, both sides were obviously meant
to be viewed, which makes it difficult to understand how the plaque was
used. That it was set into some type of frame is indicated by the upper and
lower holes.
The en face figures with triangular outlined face are of course the same as
that of the figure on the faience concave vessel also from Surkh Dum (No. 31).
Equally matched on these two objects from the same site is the crudeness
of the incised rendering of the decoration. And while no exact parallels
to match the two juxtaposed figures functioning as masters of animals are
available, there can be no hesitation in recognizing the Luristan style of the
figures. Bull men (or figures with bull horns) are common there, as is the
en face position and the flaring skirt on other figures.113 The plaque, then, is
clearly a product of a Luristan atelier (at least the incised scene, if not the
original cutting and carving of the guilloche: cf. No. 31), and thus documents
ivory working in Luristan in the early centuries of the 1st millennium bc.
No. 26. Animal-headed pin. 43.102.26; Surkh Dum 1254. Bone; L. 5.5 cm.
113 Godard 1962, op. cit. (in note 21) figs. 3436, 38, pls. 16, 18, 21, 23; Moorey 1975, op. cit. (in
note 21) 21, 24, figs. 1, 5, pls. I, IIb, IVd, the latter with an outlined face; Clercq-Fobe, op. cit. (in
note 21) nos. 50, 56.
Represented in the round is a recumbent winged equid, probably a horse,

decorated with incised designs. Its simply rendered head is small in propor-
tion to its body, and there is a collar (?) on its neck. The wings are decorated
with both a zig-zag and a vertical pattern. Circles with a central dot form the
eye and are placed on the neck and body; the same motif with the addition of
an outer rayed circle adorns the thigh. The rear end is flat with a depression
containing a hole that once held a pin. When found, this piece was encrusted
to a black and white veined stone disc, drilled through its shorter axis. Van
Loon refers to the existence of these pins from the earliest level, but it is
not known how many were recovered.114
In form and typology these pins, with the recumbent animal made of one
material and joined to a pin made of another, are the same as the many
bronze animals, mostly equids, joined to an iron pin that are reported from
Luristan, and to the well-known Hasanlu pins.115 The pin type is clearly one at
home throughout western Iran, but if the pin was made in Luristan, which
seems almost certain, it is another example of a local school of bone and
ivory carvers in that area. What is more, it speaks to a common knowledge
over a large area of nw and western Iran of objects with a common form and
function, which could hardly be called fortuitous.
114 Van Loon 1967, op. cit. (in note 6); Schmidt, op. cit. (in note 1) 211 for reference to bone
wands.
115 Moorey 1971, op. cit. (in note 6) 196197, nos. 324327; Porada 1965, op. cit. (in note 28)
166, fig. 67, pl. 29; idem, 1975, op. cit. (in note 24) 393394, no. 310.
432 chapter twelve
No. 27. Handle (7).43.102.27; Surkh Dum 824. Ivory (7); L. 7.6 cm.
This badly corroded and pitted object is pierced at one end to a depth of
about 1.7cm. One can barely make out the incised decoration which consists
of rows of lines framing bands of zigzag patterns at each end; a clear space
of about 1.3cm. separates the decoration.
The hole at one end suggests that the object was a handle. What it held
may only be conjectured, but it was probably a perishable material, since
nothing was found in the hole.
Nos. 2830. Lion figurines. 43.102.28, 29, 30; Surkh Dum 666d. Ivory; L. 2.5, 2.5,
2.9cm.
Although No. 30 is slightly longer than the others, all are the same in all
details. Each figurine represents a recumbent lion with its head resting on
its paws. The thighs, head, and ears are in relief while the mouth is formed
by two grooves. All have flat bases pierced with two holes; No. 30 has the
holes piercing the whole figure. Few distinct stylistic features are present.
At Nirnrud and Hasanlu116 small recumbent ivory calves pierced at their
bases were used as handles or grips on the lids of ivory pyxides and it is
probable that our three lions had the same function.
No. 31. Faience vessel(s). 43.102.45a, b, c; Surkh Dum 14. Faience; greatest H. of
preserved area: 12.6 cm.; D. of rim: ca. 10.6cm.
116 Max Mallowan, Nimrud and its Remains I (London 1966) 219220, figs. 173, 174; Oscar
White Muscarella, The Catalogue of Ivories from Hasanlu, Iran (Philadelphia 1980) 195196,
nos. 242245.
434 chapter twelve
Only part of this vessel is preserved, including about one-half of the rim
and a section both below and to the viewers left of the protome. The upper,
rim area is decorated with a wide border consisting of a central outlined
band of guillochesa center enclosed by an S-shaped motifframed by
two bands of vertical lines. Below is the main decorative scene depicting
a central figure flanked by large birds moving away. The central figure, en
face, has a triangular, outlined face, oval eyes lidded at the bottom only, a
rectangular nose anda slit mouth; ears are unnaturally placed at the top of
the head and frame vertical hair lines. No mustache is indicated but random
incisions on the face suggest a beard. Nothing else remains of this figure,
but he is clearly represented in the master-of-animal position. To his right
is a creature identified as a large bird by his beak and one wing, awkwardly
placed before him. The bird has a crest and feathers at its back rendered
by short incisions. Only the back feathers remain of the bird at left. All these
figures are incised in a crude and cursory manner compared to the relatively
neat rim decoration and protome.
The protome projects from the decorated area of the vessel and consists of
a bearded male wearing a bulbous hat decorated with vertical lines and two
small upright horns at the front center. His face is human with a distinctly
large nose in a straight line from the forehead, a thick-lipped mouth, human
ears, and eyes that appear to be closed, the lids meeting at the center; if
there is meant to be a mustache it is not evident. Two small legs ending in
cloven hooves join just under the beard. The figure is clearly a demon, a bull
man. Below the protome is an incised curved border (?) and another incised
object that may be the wings of a bird, now missing. In a few areas one may
still see the remains of a glaze.
The curvature of the vessel as restored may be too flared as there is very
little evidence for a flare on the preserved part; the restoration is apparently
based on comparative vessels.
Two glazed fragments, one incised with the winged area of a bird or crea-
ture (b), the other a rim fragment (c), came with the other more complete
fragments (a). While (b) might belong to the vesselits incisions are not
noticably different in execution, (c) clearly does not. Although it has the
same rim decoration pattern and band sizes, the guilloche is not the same,
neither in the size of the central circle nor in the S-curve pattern: we there-
fore must consider that fragment (c) at least represents another similar
shaped vessel.
436 chapter twelve
With regard to the technique of manufacture, one has the impression,

based primarily on the difference of quality of execution, that the vessel was
first made, apparently in the mould, and that the incised decoration below
the rim area was subsequently added.
As fortunate as we are to have this excavated vessel for study we are
equally fortunate to have good parallel pieces available, all of which have
also been excavated, and in two distinct cultural areas of Iran. First, there
is at least one other example of a decorated concave vessel from Surkh
Dum.117 On this vessel the rim has the same motifs and arrangement as the
present example, while the base has incised triangles and the main scene
depicts a lion griffin of the same type as oursnote the beakalbeit more
neatly rendered. No protome is illustrated, so until the final publication
occurs we will not know whether or not one was originally there but has
broken away. Another concave vessel comes from Susa.118 It has the same
rim pattern as ours but with a herringbone design in the lower band and
the base has slightly oblique lines. The main scene is elaborate and depicts
two well-drawn rampant bulls flanking a stylized tree; at the rim is a single
protome formed of the head and chest of an animal, either a horse or a
bull.
Two other excavated vessels, while of different shape, are demonstrably
of the same class as the concave vessels. At Kharkai in Luristan, Vanden
Berghe119 excavated in a tomb a square pyxis decorated with rosettes on all
sides. The rim decoration is the same as that on the concave Susa vessel,
except that the guilloche is more elaborate, and the base is the same as that
on the Teheran Surkh Dum concave vessel. Two protomes, identical female
heads, are placed on opposite sides of the rim; they are pierced for holding
the pegs of the now missing lid. And from Susa again120 comes a square pyxis
like that from Kharkai in shape that also has two female protome heads
like those on the Kharkai pyxis. Furthermore, the rim is decorated exactly
like both Surkh Dum concave vessels, and the main decoration has neatly
executed sphinxes and lion griffins, the latter being the same creatures on
both Surkh Dum vessels.
117 Amiet 1976, op. cit. (in note 1) 60, fig. 39 (in Teheran).
118 Pierre Arniet, Elam (Auvers-sur-Oise 1966) 500501, no. 376; H. 20.5cm.
119 Vanden Berghe 1973, op. cit. (in note 44: Archeologia) 28 and color plate.
120 Amiet 1966, op. cit. (in note 118) 498499, no. 375; Porada 1965, op. cit. (in note 28) 72,
fig. 46.
One more vessel warrants inclusion in our discussion. This is a round

pyxis from Susa121 with two pierced lugs (not protomes, but set in the same
position) and a lid, the same type that must have been associated with the
other square pyxides from Susa and Kharkai. The round pyxis has a different
decorative scheme than hitherto encountered but the vessel clearly fits into
the class under discussion. Further evidence that this vessel belongs to the
same class is demonstrated by an unpublished plain faience vessel from
Surkh Dum in the form of a round pyxis with three lugs for holding a lid
and with the interior divided into three compartments. To summarize, we
may conclude that the six vessels brought forth (seven, if fragment (c) is
indeed from another vessel at Surkh Dum), from Surkh Dum and Kharkai in
Luristan and Susa in Elam, are interrelated as one class and share more or
less certain features in common: material, rim and base design, main scene
decoration, protomes or lugs, and overlapping shapes.
Porada, discussing only the vessels from Susa, as the others were not
available to her, concluded that the faience vessels are Elamite products,
reflecting only in part (basically the shape of the square pyxides) western
influence.122 That the vessels are indeed Elamite has been neatly demon-
strated by Amiet who in 1967 published a series of enamelled terracotta
knobs associated with wall tiles excavated at Susa. The knobs were formed
as protomes of the foreparts of bulls, bull men, horses, human heads, and
seated monkeys.123 Of particular interest for our present purposes is the knob
protome of the complete bull man, which in all details and in style is the
same as the protome on the Metropolitan Museums Surkh Dum vessel: bul-
bous hat with horns, long, full beard with a horizontal base, thick lips, lidded
eyes, and the tucked-in bulls legs. What is more, the protome on the Susa
concave vessel is formally paralleled by the knob bull protome, as noted by
Amiet, while the female heads on the Susa and Kharkai square vessels are
paralleled by the human-headed knobs. Dated cogently by Amiet to the time
of Shutruk-Nahunte II, late 8th century bc, the Susa knob protomes are from
the very same period assigned to the vessels by Moorey.124
Inasmuch as all the vessels under review have been excavated in Luristan
and Elamnot merely attributed therewe have in hand a firm, not a
putative, indication of cultural exchanges between the two areas: and for
121 Amiet 1966, ibid., 495, no. 372; Porada 1965, ibid., 72, fig. 45.
122 Porada 1965, ibid., 70.
123 Pierre Amiet, lments maills du Dcor architectural No-lamite, Syria XLIV
(1967) 2746, figs. 2, 3. 513, pls. V, VI.

124 Moorey 1975, op. cit. (in note 21) 19, 21, late 8th or early 7th century bc.
438 chapter twelve
this alone, the value of the vessels is significant. For it is of interest to

note that although several scholars have noted possible Elamite artistic
influences on the art of Luristan, they have too often discussed unexcavated
objects (e.g., rein rings, weapons) to support their conclusion; with the
faience vessels we are dealing with excavated material, concrete evidence.
The style of the scene decoration of all the vessels but one, the present
example, while at home in Neo-Elamite art as known at Susa, does not fit
into a Luristan background. At the same time, however, both the motif and
execution of the main scene on the Surkh Dum vessel under review here are
clearly at home in Luristan (cf. No. 25). To resolve the apparent conundrum
it might be suggested that the Surkh Dum vessel was imported from the
south, undecorated, except for the rim, base, and the protome, and that the
design was subsequently added in Luristan (n.b. that the aforementioned
compartmented vessel from Surkh Dum is undecorated and may equally be
an import). There is a problem, however, with this suggestion. A laboratory
analysis at the Metropolitan Museum of Art resulted in the observation that
while glaze is not consistently present, apparently because of leaching, it
does appear within some areas of the incised decoration. The occurrence of
glaze at these points is consistent with the suggestion that the design was
executed before firing, that it was an original, not a secondary, feature of the
vessel. If this is indeed the case, one is then forced to seek other explanations
for the presence of the two separate styles; perhaps an artist or artisan from
Luristan assisted in the making of the vessel in the south; or the vessel was
locally made in Luristan, modelled in all details, except for the decoration,
after the imported prototype. If the glaze was added after the vessel and the
incisions were made, the first suggestion could obtain, but we do not know
and so the other alternatives must be considered.
Cylinder Seals
by Elizabeth Williams-Forte
No. 32. Akkadian cylinder seal. 43.102.34; Surkh Dum 1124. Shell C?; H. 2.8 cm.;
D. 1.45cm.
This seal shows two groups of battling gods. To the left appear two deities
grasping both a mace and the top of the other gods crown. The second group
of two deities flank a god with arms held down with palms up. The attacking
gods grasp the central deitys crown while the god to the right smites him
with a mace.
This so-called Battle of the Gods is a scene commonly represented on seals

excavated at sites under Akkadian control in both Mesopotamia and Iran
(ca. 23342154bc). Although numerous similar examples occur in Akkadian
levels of Mesopotamian sites, the closest parallel to our seal comes from the
Iranian site of Susa.125 Both the Surkh Dum and Susa seals show two deities
reaching up to pull a horn of the others miter while tugging at a single mace
held between them. In contrast, other Akkadian battles of the gods, includ-
ing the seal of Ischpum, Ensi of Susa in the time of Manistusu (ca. 2269
2255bc),126 almost always show the mace suspended in air beneath the gods
arms as if the weapon had just been dropped by one of the combatants. The
Susa seal also provides the closest analogy for the curious horned miters
composed of three superimposed tray-like forms on the present example.
These unique stylistic features shared by the Surkh Dum and Susa seals may
suggest that they are products of an Akkadian atelier in Iran127 rather than
imports from the Mesopotamian heartland of Akkadian culture.
Although the gods on our seal bear no attributes and are thus anonymous,
deities associated with vegetation or flames in similar battles have been
identified by Amiet as divinities personifying aspects of the yearly cycle of
nature.128
125 Mesopotamia: R.M. Boehmer, Die Entwicklung der Glyptik whrend der Akkad-Zeit (Ber-
lin 1965) Abb., 318, 324 (Kish), Abb., 321, 348 (Tell Asmar). Susa: Pierre Amiet, Glypticue
susienne, Memoire de la Delgation archologique en Iran XLIII (Paris 1972) 188192, pl. 146,
no. 1550.
126 Amiet 1966, op. cit. (in note 118) no. 157.
127 See Amiet 1972, op. cit. (in note 125) 190.
128 Ibid.; also Pierre Arniet, Pour une interpretation nouvelle du Rpertoire iconogra-
phique de la glyptique d Agad, RAssyr 71 (1977) 107116.

440 chapter twelve
This Akkadian seals find-spot within the 1st millennium bc sanctuary

at Surkh Dum suggests that the seal was a highly valued heirloom worthy
of offering to the god. Of the nine Surkh Dum seals in The Metropolitan
Museum of Art, certainly four and perhaps even more are antiques saved
for many centuries and finally placed within the sanctuary. Pope, Porada,
and Van Loon129 have mentioned other early seals of Mitannian and Kas-
site origin among the Surkh Dum deposits. Precedents for the practice of
depositing seals as votive objects within sanctuaries exists at sites in both
Iran and Mesopotamia, but parallels for the placing of numerous heirloom
seals within temples are rare.130 For example, at the Iranian site of Tchoga
Zanbil, over 100 cylinder seals were found deposited in the chapels situated
at the base of the ziggurat. Of the Tchoga Zanbil seals published by Edith
Porada, only one could be considered an heirloom of the age of the present
example,131 for the remainder of the seals have been dated to the late Middle
Elamite primary occupation of that site (ca. 13001100bc).132
Ghirshman and Porada133 suggested that the Tchoga Zanbil votive seals
were made in ateliers connected to the temple and carved with scenes
reinforcing their sacral nature. Amiets observation that no impressions of
these seals are known from tablets supports this suggestion.134 The seals
seem never to have been used for sealing purposes but are purely votive
objects.
Impressions of seals very similar to the Surkh Dum heirloom seals are
known from both Iranian135 and Mesopotamian136 sites. Thus, these seals
129 Pope, apud Schmidt 1938, op. cit. (in note 1) 208, note 3; Porada 1964, op. cit. (in note 28)
17; Van Loon 1967, op. cit. (in note 6) 24.

130 Iran: Edith Porada, La glyptique, in Tchoga Zanbil IV (Paris 1970) esp. pp. 35; Mesopo-
tamia: Henri Frankfort, Stratified Cylinder Seals from the Divala Region (Chicago 1955), esp.
pp. 711; temples: ibid. 7, Table 1. For a discussion and rejection of the possibility that the
large number of Jamdat Nasr seals discovered in later layers indicate continued production
of this style, see p. 3.
131 Porada 1970, op. cit. (in note 130) 8991, no. 107 (late Akkadian to Old Babylonian
period). For the only other cylinders older than the majority of the seals discovered in the
chapels see the Mitannian seals, nos. 110111, 113.
132 Ibid. 7105, 127131.
133 Ibid.; see Roman Ghirshrnaris quote on p. 4.
134 Pierre Amiet, Glyptique lamite propos de nouveaux documents, Arts Asiatiques
XXVI (1973) 365, especially p. 22. Another example of votive seals for which no impressions
are known is cited by Frankfort 1955, op. cit. (in note 130) 1617.
135 For this Akkadian seal, No. 32, see Amiet 1966, op. cit. (in note 118) no. 157; for the Old
Babylonian seal No. 33 see Amiet 1972, op. cit. (in note 125) no. 1692; and for the Middle
Elamite seal No. 34 see ibid, nos. 20262027.
136 For No. 32 see Boehmer 1965, op. cit. (in note 125) no. 347 (Nippur); for No. 33 see
clearly were not created originally as votive objects. Perhaps the 1st millen-
nium bc inhabitants of Surkh Dum considered these seals suitable temple
offerings because of their great antiquity. The later seals (Nos. 3640) may
have been fashioned as votive objects since no impressions of seals of sim-
ilar style and iconography are known to me. This implies that they date to
the same period as the Surkh Dum sanctuary (ca. 800650 bc; see above) for
which they were intended, as was the case at Tchoga Zanbil. Indeed, as will
be discussed below, seals Nos. 3640 do exhibit characteristics of a seem-
ingly regional, Luristan, origin. Their exact dates, however, are difficult to
determine because of the paucity of comparative archaeological material
from this dark age in Iranian cultural history.
No. 33. Old Babylonian cylinder seal.137 43.102.35; Surkh Dum 786. Hematite: H.
2.3cm.; D. 1.28cm.
Louis Delaporte, Catalogue des Cylindres, Cachets, et Pierres graves de Style oriental I (Paris
1920) pl. 12:7 (Tello); for No. 34 see Edith Porada, Seal Impressions of Nuzi AASOR XXIV (New
Haven 1947) nos. 613614; and for No. 35 see Anton Moortgat, Assyrische Glyptik des 13
Jahrhunderts, ZAssyr 47 (NF 13 1941) Abb. 57, 5960.
137 Previously published by Vaughn Crawford, et al., The Metropolitan Museum of Art: Guide
to the Collection of Ancient Near Eastern Art (New York 1966) 16, fig. 26.
442 chapter twelve
Inscription
dUTU-KU5.DI Samas-dajjan(!?)
R i-la-ni servant of the gods
(uninscribed)
(translated by Dr. John Huehnergard)
An enthroned figure whose head is obliterated by a chip in the stone sits

holding a cup to the right of this presentation scene. Before this figure stands
a worshipper with hands clasped and a suppliant goddess with uplifted
hands.
This seal bears a standard Old Babylonian presentation scene and inscrip-
tion.138 The scenes date in the Isin-Larsa or early Old Babylonian period
(ca. 20001800 bc) is suggested by the suppliant goddess wearing the neck-
lace with counterbalance characteristic of this period.139 Therefore, this
Mesopotamian seal appears to be the only seal of positively non-Iranian ori-
gin among the Surkh Dum seals in The Metropolitan Museum of Art.
No. 34. Middle Elamite cylinder seal.140 43.102.39; Surkh Dum 1317. Serpentine;
H. 2.56cm.; D. 1.15cm.
A deity sits upon an animal-headed throne and holds a rod and ring. Before
him is a worshipper with an animal offering. A star is in the field between
them. To the left is a secondary scene divided into two registers: above, a
lion stalks a horned animal; below, a worshipper stands before a deity who
holds a staff. In the field surrounding them are a bird above a fish, a fly, and
a fox (?).
Worship scenes similar to the present example have been found im-
pressed on tablets at Nuzi in northern Mesopotamia and on less securely
stratified examples at Susa in Iran.141 The typically Elamite characteristics of
these scenes such as the worshippers vizor-like hair-style, the deitys crown
with outward curving horns, and his animal-headed throne have been dis-
cussed by Edith Porada.142 All seals with these stylistic features, including
138
Dr. Huehnergard points out that the empty inscription case is unusual.
A. Spycket, Un Elment de la Parure fminine a la Ire Dynastie babylonienne, RAssyr
139
XLII (1948) 8996, especially pages 9394; idem, La Desee Lama, RAssyr 54 (1960) 7384.
140 Previously published by Edith Porada, The Origin of Winnirkes Cylinder Seal, JNES 5
(1946) 257259, fig. 4; idem, 1965, op. cit. (in note 28) 47, fig. 22; idem, 1975, op. cit. (in note 24)
386, no. 297e; Amiet 1973, op. cit. (in note 134) 19, pl. 12. P.
141 Porada 1946, op. cit. (in note 140) 257259, figs. 13; Amiet 1972, op. cit. (in note 125)
nos. 20222027.
142 Porada, ibid., 257259; idem, Aspects of Elamite Art and Archaeology, Expedition 13
(1971) 2834, esp. pp. 3132.

the present example from Luristan, have been dated to the mid-2nd mil-
lennium bc by Porada on the basis of one of the Nuzi seal impressions
which bears the inscription of Winnirke, the mother of the Mitannian ruler
Tehiptilla (ca. 15th14th centuries bc).143 Amiet however, in his discussion of
the related Susa examples cites numerous similarities between this group
of seals and late Old Babylonian examples. He suggests that Winnirkes
seal might have been inscribed after its importation to Nuzi perhaps from
Elam, or, alternately, that seals of this style were popular over several cen-
turies.144
Until securely stratified examples of similar seals are uncovered by
controlled excavations, the exact date of these seals, including the present
heirloom, will remain a point of discussion. The appearance of two char-
acteristic features of these scenes on earlier Syrian seals of Middle Bronze
Age date (ca. 18001600bc), however, might support the slightly higher date
suggested by Amiet. First, the goddesses with stream-like lower bodies sup-
porting the gods animal throne on two of the Susa seals are analogous
143 Porada 1946, op. cit. (in note 140) 258; for a date ca. 1400/1300bc for this seal now see
Porada 1975, op. cit. (in note 24) 386, no. 297e.
144 Amiet 1973, op. cit. (in note 134) 1920, pl. XII.
444 chapter twelve
to those on several Syrian seals, including a seal impression from Mari

dated to the period of Zimrilim (ca. 1800bc).145 Moreover, the crown with
outward curving horns considered typical of Elamite art from at least the
early 2nd millennium bc, is distinguished here by a center composed of a
series of piled-up oval forms. Identical headgear is worn by the weather god
on several unprovenanced Syrian seals of probable 18th17th centuries bc
date.146
Connections with the north Mesopotamian, north Syrian cultural sphere
for this group of seals are thus twofold. For not only were closely analo-
gous seals found impressed on tablets at Nuzi but specific motifs of the
scenes are also paralleled on artifacts produced in these northerly regions.
Although some mid-2nd millennium bc northern, Hurrian influence in Iran,
specifically in Elam, has been adduced by Labat.147 the obscure history and
chronology of this period prohibit any definite conclusions concerning the
origin and date of this seal.
No. 35. Middle Assyrian cylinder seal.148 43.102.37; Surkh Dum 528. Greyish
chalcedony (?); H. 2.72cm.; D. 1.18cm.
On this beautifully carved seal, a hero holds two long-horned animals each
suspended by one rear leg. He wears a kilt with pendant tassels and a
helmet surmounted by a lobe-like projection. In the field to the right are
an eightpointed star and a moon crescent.
Similar heroes conquering two animals appear on seal impressions on
tablets excavated at Assur in North Mesopotamia. Three impressions dated
to the reigns of the Assyrian kings Shalmaneser I and Tukulti-Ninurta I
(ca. 12741208 bc) show a similar hero wearing a kilt with tassels and holding
145 Susa: Pierre Amiet 1972, op. cit. (in note 125) nos. 2023, 2026. Mari: idem, Notes sur
le Rpertoire iconographique de Mari l poque du Pa1ais, Syria XXXVII (1960) 215232,

fig. 1; see also idem, 1972, ibid. 259 where he notes the similarity of the Susa seals stream
goddesses with those on another Susa seal that he classifies as Old Babylonian (no. 1769).
On the probable Syrian origin of this seal see Dominique Collens review of Amiet 1972 in AfO
XXVI (19781979) 104108, esp. p. 7, no. 1769.
146 Louis Delaporte, Catalogue des Cylindres orientaux et des Cachets de la Bibliothque
Nationale (Paris 1920) no. 464; H.H. von der Osten, Ancient Oriental Seals in the Collection of
Mr. Edward T. Newell (Chicago 1934) no. 303; idem, Altorientalische Siegelsteine der Sammlung
Hans Silvius von Aulock (Uppsala 1957) no. 293. See also the comments of Amiet 1973, op. cit.
(in note 134) 19 and note 1.
147 Rene Labat, Elam c. 16001200 BC, CAH II, part 2, chapter XXIX (Cambridge 1975) 379
416, esp. pp. 380381.

148 Previously published in Crawford 1966, op. cit. (in note 137) 19, fig. 30 and Shirley
Glubok, The Art of the Lands of the Bible (New York 1963) 45.
two horned animals each suspended by one hind leg. Other impressions of
similar date provide exact parallels for the eight-pointed star and the moon
crescent that appear on the Surkh Dum example.149
The enormous curving horns of the animals, and the heros helmet sur-
mounted by a lobe-like projection, however, are unparalleled on seals pro-
duced in Assur, or on artifacts from any other region of Mesopotamia as
far as I know. Reversed animals, probably gazelles characterized by their
long lyre-shaped horns, appear on a cup in the Louvre, which is not from
stratified context, but is of probable Iranian origin on the basis of its style
and bitumen material. Dated to the early 2nd millennium bc on stylistic
grounds by Amiet, these creatures are held aloft by a bull man rather than
a human hero.150 Later Neo-Elamite artifacts including a seal and a stone
relief from Susa show similar horned animals.151 Thus, whereas no repre-
sentations of a gazelle with these distinctive undulating horns appear in
Mesopotamian art, examples are found on Iranian artifacts of varying date.
Therefore, these animals may belong to a genus of gazelle indigenous to
Iran.
149 Moortgat 1941. op. cit. (in note 136) 5088, esp. pp. 7779. Abb. 59, 60, 61.
150 Amiet 1966, op. cit. (in note 118) no. 200 A, B.
151 Pierre Amiet, La Glyptique de la fin d Elam, Arts Asiatiques XXVIII (1975) 345, no. 62
(Susa), also no. 55; Amiet 1966, op. cit. (in note 118) no. 432.
446 chapter twelve
In regard to the heros helmet, similar lobed headgear is found on only

two seals, one a 13th century bc seal impression from Assur, the second a
seal in the Foroughi collection. Through an analysis of specific features of
these seals, Edith Porada concluded that they were the products of Iranian
seal-carvers.152
Other archaeological evidence as well as historical sources provide evi-
dence of Assyrian contacts with Iran in the late 2nd millennium bc. Tomb
45 at Assur yielded a typical seal of Tchoga Zanbils 13th century bc Elamite
Elaborate style.153 Assyrian textual sources record the conquest by Tukulti-
Ninurta I (ca. 12441208bc) of several cities in the Zagros previously under
Elamite control. And Assur and Elam continuously battled over control of
southern Mesopotamia during this period.154 In view of these connections,
the appearance of Iranian seals at Assur is not surprising. Curiously, how-
ever, no seals of Middle Assyrian style have been discovered at contem-
porary Iranian sites such as Susa or Tchoga Zanbil and in that respect the
present Surkh Dum example is unique.
Later 1st millennium bc artifacts excavated in areas within the sphere of
Assyrian influence frequently show a mixture of typically Assyrian picto-
rial elements with characteristics of clearly regional inspiration.155 Cities of
the 2nd millennium bc, however, like Tell Fakhariyah in the Middle Assyrian
controlled Habur region, appear to have produced seals of standard 13th cen-
tury bc Assyrian style with few, if any, indigenous Syrian features.156 Thus, the
appearance at Surkh Dum of this seal, showing what seem to be regional Ira-
nian stylistic characteristics grafted onto the standard Middle Assyrian style
and iconography, is unprecedented. That these Iranian stylistic features are
152 Porada 1971, op. cit. (in note 142) 2834, fig. 7 (Assur) and fig. 9 (Foroughi). For a possible
example of this knobbed turban see also the badly effaced seal from Marlik, Ezat Negahban,
The Seals of Marlik Tepe, JNES 36 (1977) 81102, esp. p. 92, fig. 8.
153 Porada 1975, op. cit. (in note 24) 386, no. 297g; also see B. Parker, Cylinder Seals from
Tell al Rimah, Iraq XXXVI (1975) 2128, esp. p. 35, no. 48; and idem, Middle Assyrian Seal
Impressions from Tell al Rimah, Iraq XXXIX (1977) 257268, esp. p. 260, pl. XXVII, no. 12.
154 J. Munn-Rankin, Assyrian Military Power 13001200BC, CAH II, part 2 chapter XXV
(Cambridge 1975) 274298, esp. pp. 284285; Labat 1975, op. cit. (in note 147) 386387.
155 For a discussion of Assyrian influence see Irene Winter Perspective on the Local
Style of Hasanlu, in Mountains and Lowlands, op. cit. (in note 37) 371386, and Muscarella
1980, op. cit. (in note 116) 170, 200202, 210217, 222.
156 Helene J. Kantor, The Glyptik, in C.W. McEwan, Soundings at Tell Fakhariyah (Chicago
1958) 6985. For possible indigenous Syrian elements see p. 82, no. LIII and note also Kantors
views concerning the earliest occurrence of the winged human-headed sphinx in Assyrian art
on illiustration XI. Compare the earlier winged male sphinxes on 18th17th century bc Syrian
seals like E. Williams-Forte, Ancient Near Eastern Seals. A Selection of Stamp and Cylinder Seals
in the Collection of Mrs. Williiam H. Moore (Metropolitan Museum of Art 1976) no. 3.
unparalleled on artifacts produced in the Elamite region of Iran may suggest

that they are indigenous to Luristan, a possibility that will be confirmed only
by future excavations in the Zagros region known to have been under Middle
Assyrian control.
No. 36. Late Middle Elamite (?) cylinder seal. 43.102.36.; Surkh Dum 1461.
Unglazed yellowish faience (?) with apparent metal stain; H. 3.3 cm.; D.
0.92cm.
A border of ladder pattern frames the scene. Two longskirted attendants

reach toward a table before a seated (?) banqueter. Above the table are a
vessel below an animal.
This seals material, unglazed faience, and scene, a banquet, are charac-
teristic of Middle Elamite seals (ca. 13th12th centuries bc) excavated in
Elam at the sites of Tchoga Zanbil and Susa. Although of a less angular,
abstract style than the present example, several Tchoga Zanbil seals show
more than one attendant serving a seated banqueter. Vessels and animals
frequently appear above the table.157
157 Porada 1970, op. cit. (in note 130) figs. 7476, 7980; Amiet 1972, op. cit. (in note 125) 265,
nos. 20552063.
448 chapter twelve
The ladder-patterned border appears frequently on Susa and Tchoga

Zanbil examples of both Middle and possible Neo-Elamite date, and, in
one instance, frames figures stylistically similar to those on the present
example.158 On this Susa seal from insecure archaeological context, figures
are engraved in an analogous linear style characterized by long, stick-like
limbs. Bodies are defined by the same narrow triangular skirts ending in
a geometric pattern that echoes the seals vertically hatched border.159 The
curving plumes, top-knots, or horns that surmount the beak-like profiles
of the attendants on our seal, however, are unparalleled on contemporary
artifacts. Similar profiles and horns, however, characterize so-called ibex
demons on 4th millennium bc stamp seals from sites like Tepe Giyan in
Luristan.160 Such ibex-headed or masked beings continue in the art of Iran
into the 1st millennium bc when they appear on bronzes of Luristan type.161
The great popularity of the ibex as a motif in the Surkh Dum bronzes was
commented upon by Maurits van Loon.162 Thus, the importance of the ibex
at this site may support the identification of the attendants on our seal as
ibex-headed beings.
These horned individuals serve a bent-kneed figure seemingly suspended
in air without the support of a chair. Although the omission of such an essen-
tial feature of the composition is unusual, a similarly floating banqueter
appears on a seal from Tepe Sialk, Necropole B (ca. 9th7th centuries bc).163
The Tepe Sialk banqueters legs, however, are bent in a normal seated posi-
tion whereas our figures limbs are drawn up in a crouching posture. Only
monkeys are shown in such a squatting pose, but the upper bodies of these
simian creatures always are shown in profile and never frontally as on the
present example. Similarly floating figures with bent but not drawn up legs
and with torsos viewed frontally have been characterized as goblins by
158 For development of opinions concerning the date of the border see Amiet ibid. 273
274, nos. 2131, 2134, 2091, who dates the border primarily to the Neo-Elamite period; Porada,
ibid. 98, no. 117, 50, no. 51, states that the border might last a very long time on the basis of
the Susa seals like Amiet, ibid. nos. 21312134; Amiet 1973, op. cit. (in note 134) 2425, note 1
redates the seals published in Amiet 1972, nos. 21312134 to Middle Elamite on the basis of
the Tchoga Zanbil material.
159 Delaporte 1920, op. cit. (in note 136) pl. 33:4 (Susa).
160 G. Contenau and R. Ghirshman, Fouilles de Tep Givan (Paris 1935) pl. 38:36.
161 R.D. Barnett, Homme masqu ou dieu-ibex, Syria XLIII (1966) 259276, pls. XXIV, 1, 2,
XIX, 1, 2; for early stamp seals see pl. XX, fig. 1, etc.
163 R. Ghirshman, Fouilles de Sialk II (Paris 1939) pl. XXX, 2. For chronology see Oscar White
Muscarella, Excavations at Agrab Tepe, Iran, MMJ 8 (1973) 7071, note 14.
Edith Porada. Such creatures appear on a Foroughi collection seal dated

stylistically to Iran in the 10th9th centuries bc and on a bronze pin of
Luristan type.164
The participants in the banquet on this seal thus may be typically Iranian
hybrid creaturesa goblin as banqueter served by horned ibex-headed
demonsrather than the human figures that normally appear in such
scenes. Here creatures part-animal, part-human act as human beings, pro-
viding a novel variation on the ancient Iranian theme of animals assuming
human roles.165 Though no parallels for this type of mythological banquet
exist, the apppearance of the ibex-headed demon, common in pictorial rep-
resentations produced in the mountainous region of Luristan for millennia,
alongside the goblin, known from bronzes of Luristan type, may suggest a
regional Luristan origin for this seal. Until more exact stratigraphic evidence
is provided, this seals date must lie within the period when such unglazed
faience seals showing analogous motifs or stylistic features were produced-
the late 2nd to the early 1st millennium bc.
No. 37. Late Middle Elamite (?) cylinder seal. 43.102.32.; Surkh Dum 131.
Chalcedony with iron rust; H. 3.38cm.; D. 1.34cm.
On this seal, a pair of crosses, one placed above the other, appears in the field
alongside three figures with stick-like limbs, beak noses, and horns (?). The
first figure to the right seems to hold a weapon in its upraised hand. With
elbows jutting, the two remaining figures hold a spear with one hand while
placing the other hand on a hip.
The three figures shown in procession on this seal are unparalleled. No
exact analogies appear to exist for these flat, shallowly engraved figures,
arms akimbo, with hands defined by small drillings. Their beak-like profiles
surmounted by a bent form ending in a drilling is similar to, but less clearly
rendered than, the heads of the attendants on No. 36. The relationship
between the figures on the latter seal and a particularly Iranian ibex-headed
demon or masked being was cited above and may be applicable as well to
the present more crudely carved figures. Whether the figures on this seal
are human beings wearing ibex-horned masks or composite creatures is
impossible to determine, but both clearly are related to an Iranian cultural
164 Porada 1965, op. cit. (in note 28) 78, fig. 49 and 235, chapter VI, note 6 for the pin
reference.
165 Pierre Arniet, La Glyptique mesopotamienne archaique (Paris 1961) 42, 158, pis. 37, 38; for
animals participating in banquets on earlier seals see pl. 99, no. 1308 (Ur) and no. 1313 (Tell
Asmar).
450 chapter twelve
tradition involving the hunt or worship of the ibex. The composite ibex-
headed creature is a manifestation of myth or fable, as may be the case
on No. 36, while the masked individuals perhaps reenact the myth through
ritual. The latter possibility may be applicable to the present scene for the
absence of animal prey near the armed figures and their strange gestures
may suggest that they are involved in a ritual, perhaps a dance, prior to the
hunt.166
A date for our seal in the late 2nd millennium bc is suggested solely on
the basis of the six crosses, motifs most commonly found on Kassite and
post-Kassite artifacts (ca. 14001000bc).167
The metal-like accretions on the stone may suggest that this seal was
deposited in the Surkh Dum sanctuary alongside metal artifacts. Interest-
ingly, the only other Surkh Dum seal in The Metropolitan Museum of Art
that bears similar metallic stains is the faience seal No. 36, which also shows
166 Barnett 1966, op. cit. (in note 161) 259276 for discussion and modern ethnographical
parallels. See also Porada 1964, op. cit. (in note 28) 15, and Amiet, loco cit. (in note 165).
I neglected to discuss similar connections for a seal of Proto-Elamite date in Ladders to
Heaven, op. cit. (in note 104) 191, no. 155.
167 Thomas Beran, Die Babylonische Glyptik der Kassiten-zeit. AfO XVIII (1958) 255278,
Abb. 5, 12, 18,32.

ibex-headed creatures. The possible significance of this fact will be fully

understood only after the exact find-spots of the Surkh Dum material are
elucidated in the final report.
No. 38. Early Neo-Elamite (?) cylinder seal. 43.102.40.; Surkh Dum 103. Burned
chlorite (?); H. 2.58 cm.; D. 11.7cm.
A rampant griffin attacks a couchant winged bull on this seal. In the field
surrounding these animals are a fly (?), a fish, a star, and the lower body of a
monkey (?).
Similar griffins with inward curving wings occur on two seals from Susa
and one from Tchoga Zangil dated by Amiet to the early Neo-Elamite pe-
riod.168 A seal from Tepe Sialk, Necropole B (ca. late 9th7th centuries bc)
shows an animal in analogous rampant pose menacing winged horned
animals. And a vase from Sialk dated to the same period shows griffins that
also can be compared to the creatures on this Surkh Dum seal.169
As noted by Amiet, creatures having crescent-shaped wings with feathers
indicated on their outer edge are characteristic of artifacts produced in the
early centuries of the 1st millennium bc170 An unprovenanced quiver dated
168 Amiet 1972, op. cit. (in note 125) 273, nos. 21262127; idem, 1973, op. cit. (in note 134)
pl. XVI, no. 71 (Tchoga Zanbil).

169 Ghirshman 1939, op. cit. (in note 163) pl. XXXI, 3, LXXXV, B, D.
170 Amiet 1973, op. cit. (in note 134) 26.
452 chapter twelve
on stylistic grounds to ca. 900700 bc in The Metropolitan Museum of Art

shows slender bulls with analogous arched necks and saber-shaped wings.
The lower body of what may be a monkey on our seal is an example
of a typically Iranian taste, beginning as early as the Proto-Elamite period
(ca. 32002900bc), for the abbreviated rendering of animals.171 A compara-
ble omission of specific figural components may be seen in the composite
creature on the Surkh Dum pinhead No. 7.
No. 39. Early Neo-Elarnite (?) cylinder seal. 43.102.30; Surkh Dum 807. Burned
chlorite (?); H. 4cm.; D. 1.13cm.
An animal flanks a sacred tree on this tall, slender cylinder seal. Above the
animals head is a Maltese cross and behind it are a monkey above a bird.
Although iconographically related to late 2nd millennium bc post-Kassite
artifacts of Mesopotamian and Iranian origin, the scene on this seal is
stylistically similar to later 1st millennium bc artifacts produced in Iran. A
171 Porada 1965, op. cit. (in note 28) 51, fig. 30; for Proto-Elamite seals see Amiet 1961, op.
cit. (in note 165) pl. 32:516, pl. 35:550.

scene showing linearly patterned animals alongside a stylized sacred tree

is a tableau datable by its occurrence on a kudurru (a boundary stone) of
the reign of Marduk-nadin-ahhe (ca. 10981081 bc).172 Moreover, such scenes
are common on contemporary post-Kassite seals excavated in Mesopotamia
and Iran, and on rings like those excavated at Surkh Dum (No. 20).173 Also
characteristic of the decoration of these post-Kassite seals is the Maltese
cross that appears above the head of the horned creatures on the present
example.174
Unparalleled on post-Kassite seals or on bronze rings are several distin-
guishing features of the sacred tree and the animal on our seal. Most notable
for the tree are the bent and upward-pointing branches, one with spear-like
termination, and the spikey forms that hang below it. The closest analo-
gies for these motifs appear on 1st millennium bc artifacts from Iranian and
Urartian sites like Tepe Sialk and Karmir Blur, and on an unprovenanced
Iranian cylinder seal dated to the first half of the 1st millennium bc on stylis-
tic grounds.175 On this seal, branches with spikey petals identical to those
appearing on the present example grow from the sides of a tree-deity with
curved denticulated wings typical of early Neo-Elarnite seals.
Rather than winged or non-winged bulls or caprids common on post-
Kassite examples, our seal shows a wingless composite creature which may
be a horned dragon beside the tree. Instead of hooves, the animals fore-legs
terminate in curving claws. These are similar to the claws of a feline creature
on a seal in the Foroughi Collection dated by Edith Porada to the 10th9th
centuries bc. Its rear legs, however, end in forks similar to the bird shown
directly behind it on our seal.176
Curious, too, are the short spikey horns ending in knobs in combination
with the creatures long upturning tail. On Mesopotamian as well as Iranian
examples, only bulls with one forward curving horn are shown with long
uplifted tail; two horns as on this animal identify caprids having tails usually
short, but occasionally long and hanging down between the animals legs.177
172 Beran 1958, op. cit. (in note 167) 276.

173 Ibid., 274278, Abb. 2832; Amiet 1972, op. cit. (in note 125) 273, nos. 21212125; Porada
1964, op. cit. (in note 28) 1316; idem, 1965, op. cit. (in note 28) 118120, pl. 33 for a vase
fragment from Hasan1u, 9th century bc.
174 Beran 1958, op. cit. (in note 167) 276.
175 Ghirshman 1939, op. cit. (in note 163) pl. XXXI, 2; B.B. Piotrovskii, Urartu (New York 1967)
73, fig. 55; Amiet 1975, op. cit. (in note 151) 345, esp. p. 17, pl. IX. no. 71. For trees, birds, and
monkeys, see also Amiet 1972, op. cit. (in note 125) nos. 21212122 (Neo-Elamite).
176 Porada 1965, op. cit. (in note 28) 78, fig. 49.
177 For bulls with long tails see Beran 1958, op. cit. (in note 167) Abb. 16, 28, 31; caprids with
short tails, Abb. 2224; caprid with long tail hanging down, Abb. 30.
454 chapter twelve
The Surkh Dum creatures extremely long, only slightly arched neck with
spikey forms marking its outer curve also is unusual. Linear striations fre-
quently decorate the interior volume of animals necks on post-Kassite arti-
facts but never protrude beyond the necks outline.178 On our seal, however,
the linear strokes issue from the edge of the neck and thus may indicate the
creatures mane. Analogous short linear details, some slightly up-curving
as on the present example, decorate the neck of a short horned, leonine-
clawed, birdfooted creature on a seal of probable 1st millennium bc date in
the Yale Babylonian Collection. Identified as the horned dragon of Marduk,
two of these creatures with more elaborate curling manes appear on the
seal of the son of Shutur-Nahunte II, a Neo-Elamite ruler (ca. 7th6th cen-
turies bc).179 Thus, the short strokes stretching from our creatures horn to
the base of its neck may represent a similar mane and identify it as a horned
dragon. Moreover, the curving, rounded forms of our creatures body are
more closely related to these horned dragons than to the linearly patterned
bodies of 2nd millennium bc animals.
Horned dragons never appear alongside sacred trees on 2nd millennium
bc artifacts. Seals of the 1st millennium bc, such as the Yale seal and the
seal of Hupan-Kitin, the son of Shutur-Nahunte II discussed above, however,
show the horned dragon rampant alongside a spade, another emblem of
Marduk. On the latter seal, the spade is shown as a stylized sacred tree. Since
the creature on our seal appears to be Marduks animal attribute, the horned
dragon, perhaps the sacred tree with its spade-like branch is in some way
related to that gods triangular emblem.
Thus, this seal finds its closest stylistic analogies in artifacts from early
1st millennium bc Iran. A 1st millennium bc date and Iranian origin may
be supported by the seals material, which is the same burned chlorite of
Nos. 38 and 40, seals of probable early Neo-Elamite date. Because of the
unique nature of the scene, however, and the analogies mentioned above
to post-Kassite artifacts, the possibility cannot be excluded that this scene
represents a previously unknown regional style of the late 2nd millennium
bc that perhaps is indigenous to Luristan.
178 Ibid. Abb. 28; for later material see Porada 1965, op. cit. (in note 28) pl. 33; also see
Madeline Noveck, The Mark of Ancient Man (Brooklyn Museum 1975) no. 36.
179 Amiet 1975, op. cit. (in note 151) 1819, pl. IX, no. 67 (Yale), pl. VI, no. 34 (Hupan-Kitin).
No. 40. Early Neo-Elarnite (?) cylinder seal.180 43.102.33; Surkh Dum 1299.
Burnt chlorite; H. 4.31cm.; D. 1.28cm.
The central scene shows two rampant horned animals flanking a tree with
spikey branches and tendrils. To the left an archer kneels above a horizon-
tally flying bird with denticulated wings. Above the archer is a star.
This tall, slender seal, published and discussed in depth by Edith Porada,
has been compared to several seals of different style but similar iconography
from Tchoga Zanbil.181 Of faience and late Middle Elamite date (ca. 12th11th
centuries bc), the Tchoga Zanbil seals also show kneeling archers along-
side enormous horned animals flanking vegetation. The size differential
between man and beast suggested to Porada that the caprids might be con-
sidered supernatural by the carver of the seal.182
This seal has been placed chronologically by Porada between the late
Middle Elamite Tchoga Zanbil examples (ca. 12001000bc) and the later 9th
century bc Neo-Assyrian linear style seals which more closely parallel the
180 Previously published in Porada 1964, op. cit. (in note 28) 15, pl. 1, fig. 1.
181 Ibid. 15, text fig. 1; also idem 1970, op. cit. (in note 130) pl. IV, nos. 3536.
182 Ibid. 15.
456 chapter twelve
carving of the present example. Also characteristic of early 1st millennium

bc seals of Iranian origin are the multi-rayed star and the denticulated tree
branches and bird wings on the present example.183
Conclusion (O.W.M.)
In 1977 and 1979 I presented a list of Luristan objects that have been exca-
vated both in Iran and elsewhere.184 Several more excavated pieces may
now be added, although, unfortunately, none has been published with pho-
tographs: a bronze tube with Janus heads at the top and a screw base, from
Baba Jan in eastern Luristan;185 a bronze standard finial consisting of con-
fronting felines from Xatunban in the Ilam area of western Luristan, and
from the same site five bronze horse bits, at least one of which is in the form
of winged goats trampling a gazelle.186 Collectively they add up to a total of
28 excavated Luristan objects (25 from Luristan itself and three from else-
where),187 a pathetically small number when compared to the thousands of
183 Ibid. 14.

184 Muscarella 1977, op. cit. (in note 37) 192; idem 1979, op. cit. (in note 1) 1334. For the sake
of convenience I summarize the lists here. Bronzes: three idol finials or standards (Tutalban,
Bard-i-Bal, Samos); two goat finials (Bard-i-Bal); three whetstone handles (Bard-i-Bal); one
zoomorphic headed pin (Baba Jan); one bird pendant (Bard-i-Bal); two duckheaded pins
(Kural-i-Gulgul); one open-work pendant (Crete); one bracelet with duck terminals (Bard-
i-Bal); one bracelet with animal terminals (Bardi-Bal); also a pick axe with a human face
in relief (War Kabud), and a spouted vessel with a human face at the base of the spout,
either from Luristan or a neighboring area: In addition, there are three terracotta figurines
(Chekka Sabz) and faience vessels (Surkh Dum, Karkhai).
185 C. Goff, Baba Jan, Iran VIII (1970) 176; I missed this reference because no photograph
was furnished: from the description the tube seems to be similar to those illustrated in
Moorey 1971, op. cit. (in note 6) pls. 37, 38. We thus have two Luristan bronzes from the
settlement site of Baba Jan.
186 Published too late for my 1979 paper: Exposition des denires Dcouvertes Archaeolo-
giques 19761977 (Muse Iran Bastan 1977) 42, nos. 384, 385; no. 386 mentions 1 des 4 mors en
bronze, but it is not clear if they are plain or decorated with figures. The confronting felines
are apparently like those illustrated in Moorey 1971, op. cit. (in note 6) pls. 31, 32, but we do
not know if they are of the naturalistic or the stylized type. Note also that a bronze bucket
excavated by Louis vanden Berghe, La Necropole de Charnzhi-Mumah. Archaeologia 108
(1977) 60, 61, has a scene of a city under siege and a chariot battle, known to me from a drawing
kindly sent to me by the excavator. This bucket may be an Assyrian import.
187 When the final publications of Vanden Berghe appear more objects may be added;
I have not counted. except for one example (above) the weapons excavated by Vanden
Berghe. Note also that the objects allegedly found at Maku in Iranian Azerbaijan can in no
archaeological sense be accepted as excavated objects; cf. Moorey 1971, op. cit. (in note 6) 16,
143144.
objects claimed to derive from Luristan and given much prominence in pub-
lications for the last 50 years. It is against this background that the present
report should be viewed, for it increases by over 100 % the number of exca-
vated Luristan objects available for study and discussion. Equally important,
a variety of objects not recognized among the recorded excavated types
may now be included within the bona fide repertory of Luristan artifacts,
e.g., sheet-metal work, anthropomorphic-headed pins, pendants, and ivory
and bone material. And, finally, the cylinder seals significantly amplify our
knowledge concerning the foreign relations of the Luristan culture, knowl-
edge based not on dealer-derived material, but rather on the only evidence
that can lead to meaningful archaeological conclusions-excavated objects.
chapter thirteen
NORTH-WESTERN IRAN: BRONZE AGE TO IRON AGE*
Abstract: In a paper published in 1977 and a monograph published in 1982, I.N.

Medvedskaya attempts to demonstrate that the Iron Age I period in North-western
Iran was not, as has been argued by a number of archaeologists, a break in the cul-
ture of the area. Medvedskaya asserts that not only is there evidence that ceramics
and burial customs of the preceding Bronze Age continued into the Iron Age, but
that there is no uniform culture manifest within that Iron Age I period. Each of these
assertions is confronted with the actual evidence from excavated sites in the north-
west, from Hasanlu, Dinkha Tepe, Geoy Tepe, Kordlar and Hajji Firuz. The evidence
easily refutes Medvedskayas conclusions; it demonstrates that, as previously per-
ceived, there is a major cultural break between the Bronze and Iron Ages in the
north-west, and that there is a uniformity in ceramic forms in the latter period. Fur-
ther, the author separates the issues of archaeological cultures from that of linguistic
theories concerning the language spoken in the north, issues not clearly articulated
by Medvedskaya.
By the late-9thearly-8th centuries bc the kingdom of Urartu comprised
an area extending from Eastern Anatolia to North-western Iran (Fig. 1),
with a uniform culture expressed in architecture, pottery and iconography.
Before the late-9th century bc, however, the two areas experienced different
archaeological histories.
In Eastern Anatolia, Urartian sites have revealed no material earlier than
the 9th century bc and there is little material available for study after the
period of the Trans-Caucasian culture of the Third Millennium BC (Burney
1972: 86, 127; Edwards 1986: 73). This is puzzling because from the 13th
century bc and continuing through to the 9th, that is to say, the time of the
Iron Age in North-western Iran, Assyrian texts refer to the land of U-ru-at-ri
or the Nairi lands, surely signifying that some areas of Eastern Anatolia were
occupied before the 9th century.
The apparent cultural gap in the Second Millennium in the Van area even-
tually may be partly closed if the polychrome and painted wares attributed
there could be archaeologically located with certainty. These wares are in
* This chapter originally appeared as North-western Iran: Bronze Age to Iron Age,
in Anatolian Iron Ages III, eds. A. ilingiroglu and D. French (Ankara: British Institute of
Archaeology at Ankara, 1994), 139155.
460 chapter thirteen
some cases very similar to the painted Urmia wares from the Urmia region
(Edwards 1986: 69). But when one encounters painted vessels with the very
same decoration that were purchased as far apart as Adana and Van (respec-
tively, ilingiroglu 1984: 130, 137, Figs. 1, 13 and 132, Figs. 58), one must be
hesitant about assigning provenances.1 These vessels most probably derived
from somewhere in Turkey and, as such, inform us of cultural relations, even
a common ceramic tradition, with North-western Iran sometime prior to the
15th14th centuries bc.2
In North-western Iran, excavations at Hasanlu and Dinkha Tepe in the
Solduz and Ushnu valleys respectively in Azerbaijan have revealed a strati-
graphic sequence from the early-Second through most of the First Millen-
nium bc. At both sites there exists the same Bronze Age culture (Hasanlu VI,
Dinkha IV) defined by painted, incised and plain buff wares that are related
to the Syrian and Mesopotamian Khabur wares (Fig. 2, nos. 3, 6, 1114 and
Fig. 3). This culture terminated sometime in the 17th16th century bc (Dys-
on and Muscarella 1989: 23 n. 25; Stein 1984: 29). At Dinkha polychrome ware
(Pl.1.1) occurs primarily in the Bronze Age erosion levels and in the over-
lying basal Iron Age fill, with no firmly associated architecture. Although
not yet fully studied, this ware seems to represent a post-Khabur period
occurrence that briefly existed just before and at the very beginning of the
new Iron Age period (see also Edwards 1986: 63). Further evidence for this
brief overlap are two polychrome vessels, one each from Early Iron Age I
burials at Hasanlu (Stein 1940: Fig. 110, Pls 24:3, 31:8) and Dinkha (Fig. 4.1;
Muscarella 1974: 39, 48, Figs. 3, 5); these could be imports from the north or
have been locally acquired.3 A similar overlap is also evident from painted
vessels present with Iron Age pottery at Kordlar on the western shore of
Lake Urmia (Lippert 1979: Figs. 911), at Sialk (Ghirshman 1939: Pls 37, 40, 41,
S444, S476, S495; Young 1965: 70, 72)4 and, intriguingly, at Kizilvank across
1 Modern trade has surely passed the vessels far from their plundered homes. Edwards
(1986: 70) says archaeologists believe they came from north of Van.
2 The ante quem date would be determined by this beginning date for the EWGW/Iron I
culture. Edwards (1986: 69) seems to accept the vessels in Adana as locally acquired, a prove-
nance which is not demonstrated. He also suggests (p. 70) that plundered painted vessels in
Munich are possibly from North-western Iran, a conclusion without archaeological value.
3 ilingiroglu (1984: 130, Fig. 2) compares a vessel in Adana with the Hasanlu example.
While close, it is not exact. Edwards (1986: 63 n. 61) incorrectly assigns the Dinkha vessel to a
stone tomb. It actually came from an inhumation cut into the Bronze Age level. He dates the
burial to the Bronze Age, with no documentation or reference to the published text.
4 Medvedskaya does not discuss these painted vessels, which may represent a degree of
continuity. In her illustrated pottery chart they are not shown in the Sialk row.
north-western iran: bronze age to iron age 461
the Araxes River (Dyson 1973: 705), where polychrome painting occurred on
Iron Age (in Iranian terms) pottery shapes.
Stratified, pre-Iron Age polychrome wares occur elsewhere in the Urmia
areaabundantly at Haftavan, at Kordlar, in the Trans-Caucasus and, as
noted, in Anatolia (ilingiroglu 1984; Edwards 1986: 64, 67, 70, 72),5 all of
which attests to a cultural tradition of painted wares linking these areas in
the Bronze Age, sometime about the middle of the Second Millennium bc.
Commencing with Hasanlu period V and Dinkha period III, the Bronze
Age cuLture was succeeded in the 15th or 14th century bc by a distinctly
new culture characterized by burnished monochrome grey and orange-red
wares (Fig. 4.2 and Pl. 1.2). The culture was first called the Early Western Grey
Ware (EWGW) horizon by Young (1965: 70; 1985: 362 n. 1), then simplified to
Iron I by Dyson (1965: 211; Young 1967: 22 n. 68), although still technologi-
cally a Bronze Age culture (on terminology, see Muscarella 1974: 79; Lippert
1979: 134: die bronzeitlichen Tradition; Haerinck 1988: 64 n. 1). The follow-
ing period, Hasanlu IV and Dinkha II, clearly a continuation of the same
culture, was called the Late Western Grey Ware (LWGW) horizon or Iron II
(Pl. 2.1; Young 1967, 1985; Dyson 1965). At Hasanlu this period was terminated
in a major destruction about 800 bc, most probably accomplished by the
Urartians, who then occupied the site (Dyson and Muscarella 1989: 3, 20;
Muscarella 1989: 34).
When the Urartians destroyed Hasanlu IV, and most probably Dinkha II,6
they destroyed a culture that had existed in the area for about 600 years
(Dyson 1965; idem 1977; Muscarella 1974; Dyson and Muscarella 1989). But
the culture that had existed in the Van area during this period remains
archaeologically unknown.
It was clear from the early excavations at Hasanlu that the Iron Age cul-
ture was distinct in a number of aspects, most obviously in pottery forms
and surface treatment from that of the earlier Bronze Age culture. This dis-
tinction was made manifest by the subsequent excavations at Dinkha, where
5 Edwards map (1968: 71, Fig. 4) excludes Dinkha from the polychrome zone. For painted
and polychrome wares from pre-Iron Age Kordlar (not discussed by Edwards 1986), see
Kromer and Lippert (1976: 6582, Pls 46) and Lippert (1976: Pls 1, 2). Also compare Kromer
and Lippert (1976: Pl. 4.2) with ilingiroglu (1984: 130, 137, Figs. 1, 13 and Pl. 12.1) here with
ilingiroglu (1984: 131, 135, 136, Figs. 2, 9, 12).
6 We have no evidence at Dinkha for or against a destruction, as the Iron II buildings
preserve only one course of foundation stone, but no postLWGW/Iron II material exists
there. Dinkha lies just across the valley and in sight of the Urartian site of Qalatgah built by
Ishpuini and Menua, and surely would not have been left free (Muscarella 1974: 58, 82; 1989:
34).
stratified architecture, burials and pottery of the Bronze Age was available
for study alongside the later EWGW/Iron I finds there (similar but quanti-
tatively less evidence is also now known from Hasanlu). That this distinc-
tion indicates a major break and a new beginning in the history of the area
has been recognized by all scholars who have reviewed the evidence (e.g.
Dyson 1965: 197; 1977: 156, 166; Young 1965: 57; 1967: 22, 24, 31; 1985: 361, 364,
373; Burney 1972: 115, 117; Muscarella 1974: 52, 81; Levine 1987: 233). Further
research demonstrated that the same EWGW/Iron I horizon is present at
other sites surveyed and excavated in the Solduz valley (e.g. Hajji Firuz,
Nagadeh); on the western shore of Lake Urmia, at Kordlar and Haftavan;
in the Elburz Mountains near Teheran, at Khurvin and Gheytaryeh (which
have not yielded earlier material); at Sialk; and in limited form at Godin and
Giyan in Central-western Iran but not in the Mahidasht area or in Luristan
(Young 1985: 364; Levine 1987: 241, 247). The EWGW/Iron I horizon was char-
acteristic in Iran only in the north-west and along the southern face of the
Elburz mountains (Young 1985: 367, 374): an essential understanding neces-
sary to keep in mind in discussions of the Iron Age(s) in Iran at large (Young
1985: 363 n. 1).
Furthermore, that this new culture, appearing in the area a century or
more after the termination of the Bronze Age Khabur culture,7 indicates
the presence of a new ethnic entity, seemed to most scholars a reasonable,
albeit not provable, conclusion. However, to go beyond the archaeological
evidence and attempt to identify the ethnic affiliation of the newcomers
is another matter. Thus the assertion that the newcomers were Iranian
speakers (e.g. vanden Berghe 1964: 46; Young 1867, 1985: 372, 374some
Iranians; Burney 1972: 115; Ghirshman 1977: 47, 50, 54, 62) is speculative,
being incapable of documentation, since written records have yet to be
recovered.8 And opinions regarding the direction and loci from which this
Iron I culture arrived are still under review and are not easy to resolve
(compare Young 1967 and 1985 with Dyson 1977).
Nevertheless, taking a position that may accept or reject a specific ethnic
identification or a preference for one route over another, one may admit that
7 At Dinkha in the B9 and 10 areas, along the north, there is evidence of a chronological
gap between the end (by burning) of the Bronze Age and the incipience of the trash deposits
into which was created the Iron Age cemetery (Muscarella 1974: 52). The excavations at
Hasanlu were too limited for this period to allow for a firm conclusion about a gap (Dyson
1965: 195).
8 Young (1967: 31) also thinks that Marlik (which has metal spouted vessels similar to
Iron I shapes) was an Iranian (language) site. Medvedskaya (1982: 11) also has the same view,
which seems a contradiction to her general conclusions (infra).
there are issues separate from that which accepts the conclusion that a new
culture (most probably brought by a new people) entered North-western
Iran, which is strictly an archaeological issue.9 And this latter conclusion is
a basic fact of the archaeological record.
In a paper (1977) and monograph (1982) I. Medvedskaya is convinced that
she has succeeded in demonstrating that there was no cultural break from
the Bronze to the Iron Age in North-western Iran, that there was a continuity
in the culture. Furthermore, in as much as there was continuity, there can be
no argument that Iranians arrived there in the Iron I period; indeed, there
can be no claim that any new ethnic group arrived there at this time (1977:
104; 1982: 38, 96, 99, passim).10
To demonstrate these historically significant conclusions, Medvedskaya
claims that a careful reexamination of the finds and their contexts indicates
that 1) there is a lack of cultural unity in the Iron I period since it is not stable
(1977: 94, 99; 1982: 13, 34, 39, 50, 96) in pottery distribution, that 2) Iron I
pottery forms are not new for Iran in the period Iron I for they contain
Bronze Age forms (1977: 103; 1982: 34, 38, 98) and consequently the origin
of the pottery is to be sought in situ in the roots of the local (Bronze Age)
culture or the cultures of adjacent areas, and that 3) burial forms are not
distinct in the two periods.
In all these assertions Medvedskaya is utterly wrong. An analysis of her
presentation of the evidence reveals that in almost every instance there is
misunderstanding of the published material, omission of relevant and sig-
nificant information and errors of fact. What is unfortunate is that her con-
clusions, unexamined, are being cited by scholars either as a new and correct
interpretation of the early history of Iran (Genito 1986: 50; Dandamayev and
Lukonin 1989: 12) or as useful (Curvers 1985: 198) or casually, as a different
view (Henrickson 19831984: 216 n. 51). Lest still others uncritically cite the
subjective revisionism of Medvedskaya, that is, cite her without first going
back to the data, I will here review her claim.11
9 Note that neither Dyson nor the author ever stated that the EWGW/Iron I culture
represented Iranian speakers. Dyson at one time (e.g. 1960: 119, 121; 1964b: 3) thought the
entity in Hasanlu IV (LWGW/Iron II) was that of the Manneans, a position later dropped;
see Boehmer (1964: 19 n. 64), who rejects Hasanlu IV as a Mannean site.
10 Since Medvedskaya (1982: 1) believes that the Urmia area was occupied by Hurrians in
the Second Millennium bc, it would follow that she must also believe (a belief avoided in the
text) that the inhabitants of Hasanlu and Dinkha were Hurrian speakers in the Iron I and II
periods.
11 The text of the monograph is often difficult to understand. Information on specific sites
is not given in one place but scattered. The same scattering exists in the pottery and burial
First the claim of pottery continuity. To document her charge, Medved-

skaya spends much time on the istikhan, a typical Bronze Age cup, and she
utilizes it as a crucial argument for continuity. Eleven plus examples are
cited as deriving from the Iron I period (1982: 17, 27, 36), including Hasanlu
(32, Fig. 4, Row 4; here Fig. 2 nos. 13, 14) but in reality there is only one exam-
ple known from this period, from Sialk A (Ghirshman 1939: Pl. 40, S472).
All the others cited are either not true istikhans or they are from undated
(Haftavan)12 or Bronze Age sites (Godin) or from undated survey material
(Bijar). Two early publications (Dyson 1964a: 36; Young 1965: 67, Figs. 8, 11)
that reported istikhans from a single burial attributed to Hasanlu V were
wrong: the burial belongs to Hasanlu VI, a fact corrected by Muscarella (1968:
195) and Hamlin (1974: 148 no. 13).13 The Hasanlu istikhans probably derived
from the west, as Medvedskaya suggests (1977: 100; 1982: 36), but as part of
the Khabur corpus: which speaks for western relations in the Bronze Age,
not the Iron Age.
Medvedskaya (1977: 84, 101; 1982: 36) then refers to painted pots with
bands and bowls with painting on the rim from Hasanlu V, material which
she also utilizes to deny uniqueness to the Iron Age pottery there, and again
to record derivation from the west. She is right with regard to western ori-
entation but again wrong with regard to the correct period of this orienta-
tion, for there are no examples of these vessels from Hasanlu V: all are from
Hasanlu VI (Fig. 2 nos. 3, 6) and no such vessels occur in Dinkha III. To be
fair, this is the only instance where Medvedskayas conclusions can be said
to have been misinformed by early publications (Young 1965: 70, 72 and the
mislabelled Fig. 8:3, 6, Fig. 11; now see Young 1985: 373 n. 11).14 For examples
of two painted vessels of a different type that do derive from early Iron I
contexts, see Fig. 4.1 here and supra.
discussions. Sometimes the meaning of a sentence is unclear: this may be a problem with the
translation.
12 Burney (1973: 164) dated istikhans to the Iron I period solely because he thought
(wrongly) that the Godin examples were from this period; but note that Burneys text here is
unclear: were true istikhans found at Haftavan?
13 These corrections are not reported by Medvedskaya although the works are cited in her
bibliography. Medvedskaya also omits the istikhan found by Stein (1940: Pl. 30: 11) in an Iron II
burial. There are two metal examples with handles from Tepe Guran (Meldgeard, Mortensen
and Thrane 1964: 128, Fig. 29). The published Hasanlu istikhans derived from a period VI
(Bronze Age) burial, OP 10, burial 2 (Dyson 1959: 9). Other istikhans occur in the period VI
level at Hasanlu.
14 Medvedskaya (1982: 19, 1(b)2) (she gives no reference and refers to grey and cream
colour) may be referring to the one example from Dinkha III but she does not mention the
one from Hasanlu (supra). Medvedskaya (1977: 94) mentions painted wares in significant
numbers in Hasanlu IV: there are no painted wares from this period.
Medvedskaya (1977: 96; 1982: 13, 39) claims that pattern-burnished pottery
occurs often and is a decoration characteristic of Iron I pottery but (Dyson
1964a: 36) specifically stated that it was rare in that period. Furthermore, she
finds it possible to cite a unique pattern burnished sherd from Dinkha IV
(Hamlin 1974: 129) as an indication of continuity, even though there is no
example known from Dinkha III (Muscarella 1974: 48).
Medyedskaya (1982: 38,98) notes that grey wares are ubiquitous in the
Near East in the Second Millennium bc, in Mesopotamia, the Caucasus,
Greece and Iran, including Dinkha IV, all of which attests to the continuous
existence of grey and also grey-black ware throughout the Second Millen-
nium bc. Therefore, the sharp increase in its quantity in the Iron Age I can-
not be considered sudden and, what is more important, it was not brought
from outside. What is not realized in this argument is that grey ware is here
treated generically, grey ware equals grey ware. Notwithstanding this facile
generalization, there is a difference between the grey wares of the Bronze
Age and that of the EWGW/Iron I horizon in fabric, technique of manufac-
ture, surface treatment and shape.15 No one who has closely worked with
the pottery in Iran would confuse the two (Young 1985: 373: 373, n. 11; Dyson,
personal communication). Indeed, surveys in the Solduz valley have allowed
several Iron Age sites to be recognized specifically because of the character-
istic grey wares present. The grey wares of the EWGW/Iron I horizon were
brought from outside as a coherent assemblage. There is no evidence for any
in situ evolution.
Medvedskaya further posits that there was no ceramic horizon in the
corpus of the Iron I period in Western Iran (1977: 95; 1982: 50, 96). In the
north-west the lack of unity is indicated because all the diagnostic shapes
do not occur together in the same tomb or because they do not always
occur together at the same time. Consequently if every tomb does not
contain all the diagnostic shapes or if they do not occur together at each
site, cultural unity is lacking, And the pottery of the other areas in Western
IranLuristan, Southern Caspianis different.
Medvedskaya does not include a definition or explanation concerning
the reasons which determined the utilization of other geographical and
cultural areas in her study of North-western Iran and she omits significant
information from the north-west that witnesses against her views. Thus, in
the first place, to document diversity she gives equal weight to the finds from
Hasanlu V and Dinkha III in the north-west, both classical EWGW/Iron I
15 Hamlin (1974: 128) notes that the grey wares from Dinkha IV consist mainly of bowls.
sites and to those of Giyan 116 and Godin to the south-east (1982: 33, 39,
41, 54, 64). Now, while participating in some manner, and being somewhat
contemporary, with the EWGW/Iron I culture (Young 1965: 62; 1985: 366),
the southern sites cannot be brought forth to document a lack of unity in
the north. Godin preserves only three graves. The culture of Giyan both in
the Bronze II and Iron I Age periodswhich exhibit a lack of continuity
was essentially peripheral to that in the north-west during both periods
(Henrickson 19831984: 210 n. 29, 215). The neighbours valleys (including
Luristan) also lack evidence of extensive occupation of the EWGW tradition
(Young 1985: 367; Medvedskaya 1982: 96!).
Of course there was diversity in the cultures of Western Iran in the gen-
eral, chronological Iron I period, a fact recognized by all scholars who have
studied that period (e.g. Levine 1987: 232, 243; Young 1965: 362 n. 1, 364;
Henrickson 19831984: 216 n. 51; Haerinck 1988: 64, 67). None of these schol-
ars has challenged unity in a specific area. The unity of the EWGW/Iron I
horizon in the northwest cannot be refuted either by reference to other, dis-
similar, Iron I cultures or by omitting information from the relevant area
that indeed attests to unity. This is the methodology of Medvedskaya.
Medvedskayas analysis and statistics are significantly compromised by
omitting from the analyses the pottery from the important northern site of
Kordlar, which with Hasanlu and Dinkha is a typical EWGW/Iron I site, and
that from the burial at Hajji Firuz (Pl. 2.2; infra), which is most probably a
deposition from one of the several Iron Age sites surveyed in the Solduz val-
ley. When the finds from these sites are placed back (because removed by
Medvedskaya) into the corpus and into the discussion, they play a signifi-
cant role in the collapse of Medvedskayas conclusions.
As has long been recognized, there are three basic diagnostic shapes
for the EWGW/Iron I horizon: 1) the bridge less spouted vessel, 2) the
pedestal-based, goblet with handle and 3) a bowl with a small modelled
ridge in crescent form on the inside, the so-called worm bowl (Fig. 4.2 and
Pl. 1.2); also illustrated by Muscarella (1968: 193, Fig. 17).17 In her eagerness
16 It may be stated that she considers Giyan to be more important in her argument than
Dinkha, for inexplicably in her Fig. 4, an illustrated chart showing ceramic links between
the Iron I sites of Iran, there is a Giyan column but none for Dinkha, which has the largest
and most comprehensive corpus of EWGW/Iron I pottery published to date. This is but one
example of the distortion of her analysis.
17 Medvedskaya (1982: 37, 40, 42) gives much attention to cups, which she claims are
diagnostic. For cups from this period, add to her listing of Sialk and Khurvin (= Young 1965:
Fig. 9.2), examples from Hasanlu, Haftavan and Gheytaryeh, all in North-western Iran. For
Fig. 12.4.2, the B 10b, burial 13, see Muscarella (1974: 85, the catalogue).
to deny cultural unity, Medvedskaya does not mention (or is unaware of)
a fact essential to the discussions (and it is one that readily contradicts her
conception), namely, that all three diagnostic shapes are associated together
at six sites in North-western Iran: Geoy Tepe (Dyson 1965: Fig. 2; here Fig. 5),
Haftavan (Burney 1970: 165), Hajji Firuz (Muscarella 1974: 49; here Pl. 2.2),
Kordlar (Lippert 1979: Pls 5, 6), as well as at Hasanlu and Dinkha Tepe
(Fig. 4.1, 2 and Pl. 2; Muscarella 1974: 48). Furthermore, at Geoy Tepe and
Hajji Firuz18 the three diagnostic shapes occur together in one tomb and at
Kordlar they were all found together in one room.
It should also be noted in the discussion that Medvedskaya is unaware of
the importance of the worm bowl as a diagnostic shape in the EWGW/Iron I
corpus; it is found at six sites, not three as recorded in scattered references
by Medvedskaya, and it occurs only in the core area of the EWGW/Iron I
culture.19
Medvedskaya believes that she has demonstrated the lack of unity and
stability in the EWGW/Iron I horizon by asserting that its diagnostic pottery
shapes never existed side by side in a horizon. To her, they were manufac-
tured at different times and accordingly arrived in the north at different
times and from different areas.20 Aside from ignoring the evidence of the
Geoy Tepe and Hajji Firuz finds, these assertions are presented without anal-
ysis of stratigraphical or archaeologically derived chronological evidence
that might shed light on the major and still unresolved problem of the rel-
ative dating of the various Iron I sites in Iran. For example, she notes the
predominance of the diagnostic goblet in the north (1982: 37) but asserts
without analysis or evidence that it arrived there from Luristan [sic]
(Giyan), where it appeared earlier. In fact, our knowledge of the relative
chronology of the appearance of the goblet at Giyan and the northern sites
is not sufficient to argue that it could have begun [at Giyan] either ear-
lier or later than in North-western Iran (Henrickson 19831984: 212).21 And
when Medvedskaya (1982: 50) states that the EWGW/Iron I goblet must have
come from the pottery of the previous period, she in fact cannot cite a sin-
gle earlier example.
18 Hajji Firuz is mentioned in her catalogue of sites (p. 102) but its vessel inventory is
ignored in the text.

19 On p. 22 Medvedskaya lists one from Dinkha; on p. 29 she adds Geoy Tepe; in Fig. 4, row
7, she adds Hasanlu (but it is omitted from Fig. 3!).

20 Thus, vessel shapes came to Khurvin from Sialk, from Giyan to Sialk, or moved west to
east, east to west, south to north, but never from north to south or east.
21 Furthermore the goblet form at Giyan is the only shape that links that site to the north
and it is not precisely matched in the two areas.

In Medvedskayas scheme, not only do goblets arrive late in the north,

another diagnostic shape, the spouted vessel, arrived still later (p. 38) and
thus cannot be a component of the EWGW/Iron I ceramic horizon. She
admits its uniqueness in North-western Iran during the Iron Age period
there, for indeed it does not occur in the area before that time (1982: 42,
47, 50). But because the basic (generic) shape occurs elsewhere in the Near
East (p. 47), it is not specific to Iranian [here apparently used ethnically?]
culture. Note also that two bridgeless spouts derive from the period V
occupational level on the Hasanlu citadel and spouted vessels come from
the cemetery (pace Medvedskaya 1982: 37, who claims but one spout from
the site and omits the shape from the Hasanlu V column in her Fig. 3).
Medvedskaya goes so far in her unresearched claims that she feels free
to assert (p. 38) that the goblet and the bridgeless spouted vessel cannot
belong to one culture.22
Now, although Tomb VIIB B2 at Dinkha (Pl. 3; Muscarella 1974: 38, Figs. 3,
4) is mentioned (but only once as an aside), Medvedskaya is unaware of
both its contents and chronology and consequently ignores its significant
evidence. For in addition to the claim that goblets and spouted vessels do
not belong to the same culture (supra), she further asserts (p. 38) that the
two shapes are not found together in Azerbaijan. But in the real world of
archaeology, not only does a goblet and a spouted vessel occur together, are
found together in Tomb VIIB B2, the contents of this burial are arguably
the earliest EWGW/Iron I material recorded up to the present time. These
contents inform us that from a very early time (perhaps even the earliest) in
the EWGW/Iron I horizon these two diagnostic shapes were manufactured
at the same time. They existed side by side.23 Moreover, spouted vessels exist
in other early Dinkha III burials (Muscarella 1974: 41, B9a burials 23, 24, 26, 27;
the latter also contained a worm bowl). As noted above, the spouted vessel
and goblet occur together or in the same chronological horizon at several
EWGW/Iron I sites in the north (viz. Figs. 4.2 and 5 and Pls 1.2 and 2.2).
To Medvedskaya, not only do pottery forms demonstrate continuity and
lack of uniqueness from the Bronze Age to the Iron Age in North-western
Iran but burial forms equally document continuity of culture (1982: 61, 67,
98). And just as she has both manufactured modern artifacts of information
22 Medvedskaya never notes that the EWGW/Iron I shapes (1982: Fig. 7: 110113) isolate
themselves from the forms of other spouted vessels in the Near East (1982: Fig. 7).
23 Medvedskaya (1977: 99) incorrectly states that all Dinkha III burials contained goblets;
the goblet (1982: Fig. 3) is omitted from the Dinkha III row.
and ignored ancient evidence to arrive at her ceramic conclusions, likewise

she has employed the same methodology for the burials.
It is asserted (1982: 54,62,64) that Dinkha Tepe in period IV, the Bronze
Age, yielded three forms of burials, 1) cists, 2) the pit or simple inhumations
(in two cases lined with bricks) and 3) urn burials. Since Dinkha III had
brick lined tombs, inhumations and urn burials, continuity is confirmed.
The archaeological facts as published, however, (if one leaves aside the
ubiquitous inhumation burial that has no comparative cultural value) are:
there are no urn burials at Dinkha that pre-date the LWGW/Iron II period
(Muscarella 1974: 75, Figs. 4951). Stein (1940: 373) actually recorded one, not
two, Dinkha IV inhumations, covered, not lined, with three bricks.24
But how does one confront the lack of cist tombs in Dinkha III? In as
much as at Geoy Tepe Tomb K, with EWGW/Iron I contents (Pl. 2.2), is
indeed a stone cist tomb, the absence of this form at Dinkha (she also cites
Godin, which is irrelevant to the issues under review) is puzzling to Medved-
skaya (1982: 64). She, therefore, ingenuously suggests that its absence there is
possibly a consequence of small excavations, which I take to mean incom-
plete clearing of the cemetery area. Fiat replaces archaeological evidence,
a leitmotif of the monograph. The fact is that cist tombs do not exist at
Dinkha III and Hasanlu V.
The Geoy Tepe tomb with its multiple burials is actually the sole example
of a burial form that existed both in the Bronze and in the Iron Age periods.
But what it signifies is unclear and we do not have enough information to
allow for easy interpretation. We do not know whether it indicates a one-
time Iron Age I multiple burial deposition or a sequence of burials over
time or whether in either case it could have been a re-use of a discovered
Bronze Age tomb or was built by Bronze Age survivors. These are problems
not raised by Medvedskaya. In any event, by itself this isolated cist tomb
cannot be utilized to assert a general continuity of culture in the area against
all the evidence to the contrary, both at Geoy Tepe and elsewhere.
Medvedskaya (1982: 57) then confronts the issue of extramural cemeter-
ies, which in her unexplained definition is not a separate burial area outside
the city walls but one existing some distance from the settlement, as in
Luristan and Sialk.25 To her, the cemetery at Dinkha (Muscarella 1968: 189,
24 A vessel from this burial is paralleled by another from Dinkha IV (Hamlin 1974: 133,
Fig. 1:4, 147, no. 4, from B 10a, burial).

25 Hamlin notes a parallel at Sialk but gives a wrong reference; it is a similar but not exact
parallel (Ghirshman 1939: Pls 3:1, 41. B492).

Fig. 2; idem 1974: 37, Fig. 2) is not an extramural cemetery (p. 65) because no
settlement of that period was recovered! The fact that at Dinkha the period
III cemetery has the same topography as the preceding Dinkha IV period,
i.e. it was deposited over Bronze Age burials, is a further indication to her
that the cemetery was not extramural.
Although Medvedskaya is correct that no period III settlement was recov-
ered at Dinkha (the whole mound was not cleared: Muscarella 1974: 36.
Fig. 1), manifestly a settlement did not exist in the cemetery area, a fact
which signifies in normal archaeological reasoning that the settlement and
cemetery were discrete and that the latter was extramural. Bronze Age buri-
als did exist below the Iron Age burials but Medvedskaya neglects to report
that the Bronze Age burials were associated with architecture. They were
intramural burials and thereby markedly of a different nature from the over-
lying burials in a cemetery dug into trash debris (whose depositors probably
knew nothing of the underlying architecture and burials).
In Medvedskayas unique system of classification, not only is Hasanlus
cemetery (which is situated in an area outside the settlement) not extra-
mural, it is not even a cemetery (p. 61). It is a pseudocemetery, which
unfortunate term she defines as a cemetery on the edge of a settlement.
That this edge is outside the settlement, and therefore extramural, is not
appreciated.
Medvedskaya (1982: 55, 65) also takes much time to discuss burials at
Giyan, Bronze and Iron Age Godin26 and in Luristan proper, all of which are
irrelevant to the archaeological history of the Bronze and the EWGW/Iron I
cultures of North-western Iran (supra), allegedly the topic of the paper.
Finally, there is a long, and equally irrelevant, chapter (Medvedskaya
1982: 68) on daggers, axes and arrow heads, the purpose of which is to
demonstrate that in the Iron I and II periods of Western Iran various forms
existed in different regions. There is absolutely nothing in this conclusion
that is new, nor in the claim that the Iranian27 weapons do not reveal
evidence of the presence of the Iranian nomads (p. 94). The chapter is but
another attempt to focus on the cultural diversity throughout Western Iran
during the Iron Age in order to document an alleged lack of cultural unity
in the northwest.28
26 The Godin excavations had nothing to do with the Hasanlu Project: it was an indepen-
dent venture. Its three Iron I graves were also extramural (Henrickson 19831984: 205).
27 Medvedskaya does not clarify her usage of the term Iranian for the cultures of the area
known as Iran and for its linguistic meaning. These are two different concepts!
28 Medvedskaya says nothing of the continuity of toggle pins and torques from the Bronze
Even approached in limited terms, for example as an analysis of distri-

bution and influences, this chapter is misleading and has no archaeological
value.29 It is deeply flawed by a lack of comprehension that unexcavated,
plundered objects found in museums and private collections have no prove-
nance and therefore cannot be introduced into a study of ancient distribu-
tion patterns for Iran or elsewhere. To Medvedskaya, the inscribed weapons
she discusses derive always within Luristan (p. 68),30 which observation
can only be defended by citing a crystal ball, not archaeological activity
(Muscarella 1988: 35, 39): for, nota bene, not a single example she cites was
excavated, nor has any inscribed metal objectweapon or otherwiseever
been excavated in Luristan. Other unexcavated objects are also arbitrarily
provided with a Luristan provenance and treated as archaeological data in
her discussion of distribution.
In summary, it is the firm conclusion of this paper that Medvedskaya
has not demonstrated that the EWGW/Iron I period of North-western Iran
did not represent a new culture in the area. A review of her assertions and
the alleged evidence she presents to support her view all the more forces
upon us the understanding that the conclusion of scholars writing on the
Iron I archaeological pattern before 1982 is still fully intact and is historically
sound.31 Indeed, if any region studied by archaeologists can be said to allow
unambiguously the conclusion that cultural change has occurred, it is in
North-western Iran in the EWGW/Iron I period.
Acknowledgements
All photographs (except Plates 1.2 and 3) and all figures (except 4.1 and 2)
are courtesy of the University Museum, University of Pennsylvania.
Age in general to the Iron Age (Muscarella 1974: 51, 52, 54, 80) nor of the polychrome wares
supra. Torques are a new element at Dinkha III. They do not appear in the Bronze Age tombs
and toggle pins all but disappear.
29 Relevant bibliography is missing, for example Boehmer and Cleuziou on arrowheads.
Medvedskaya is unaware of the literature on fibulae in the Near East and Iran: on p. 75
she dates one to the 9th century bc, when it first appears a century or more later; see also
Haerinck (1988: 65).
30 Medvedskaya uses words such as accidental or chance finds to describe what was,
and is in fact, organized plunder. Her usage distorts the magnitude of the destruction of
Iranian cultures (Muscarella 1988: 34).
31 Young (1985: 374) suggests that the people of the EWGW/Iron I culture may have
included Iranians and non-Iranians. It begs the question, How do we know. This position
seems to be an artificial compromise with his original 1967 paper.
Bibliography
vanden Berghe, L.
1964 La ncropole de Khurvin. Istanbul.
Boehmer, R.M.
1964 Volkstum und Stadte der Manner. BaM 3: 1124.
Burney, C.
1970 Excavations at Haftavan Tepe: 1968. Iran 8: 157171.
1972 Peoples of the hills. New York with D.V. Lang.
1973 Excavations at Haftavan Tepe 1971: third preliminary report. Iran 11: 153172.
ilingiroglu, A.
1984 The Second Millennium pottery tradition of the Van Lake basin. AnatSt 34:
129139.
Curvers, H.H.
1985 Review of Medvedskaya 1982. BibO 42.1/2: 197198.
Dandamaeuv M. and Lukonin, V.G.

1989 The culture and social institutions of ancient Iran. Cambridge.
Dyson, R.H. Jr.

1959 Digging in Iran: Hasanlu 1958. Expedition 1(3): 418.
1960 Hasanlu and early Iran. Archaeology 13: 118129.
1964a Notes on weapons and chronology in Northern Iran around 100 BC. In
Mellink, M. (ed.) Dark Ages and Nomads 3245. Istanbul (Netherlands Hisr.-
Arch. Inst., Publ. 18).
1964b Ninth century men in Western Iran. Archaeology 17: 311.
1965 Problems of protohistoric Iran as seen from Hasanlu. JNES 24: 193217.
1973 The archaeological evidence of the Second Millennium on the Persian
plateau. In CAR 2.1: 686715.
1977 Architecture of the Iron I period at Hasanlu in Western Iran and its implica-
tions for theories of migration on the Iranian Plateau. In Le Plateau Iranien
et lAsie centrale des origines la conqute Islamique: 155169. Paris (CNRS
no. 567).
Dyson, R.H. Jr. and Muscarella, O.W.

1989 Constructing the chronology and historical implications of Hasanlu IV. Iran
27: 127.
Edwards, M.
1986 Urmia Ware and its contribution in North-western Iran in the Second
Millennium BC. Iran 24: 5777.
Genito, B.
1986 The Medes: a reassessment of the archaeological evidence. EW 36.13; 11
81.
Ghirshman, R.
1939 Fouilles de Sialk. Paris.
1977 LIran et la migration des Indoaryene et des iraniens. Leiden.
Haerinck, E.
1988 The Iron Age in Gilanproposal for a chronology. In Curtis, J. (ed.) Bronze-
working centres of Western Asia c. 1000539BC: 6382. London.
Hamlin, C.
1974 The early-Second Millennium ceramic assemblage of Dinkha Tepe. Iran 12:
125153.
Henrickson, R.C.
19831984 Giyan I and II reconsidered. Mesopotamia 1819: 195220.
Kromer, K. and Lipert, A.

1976 Die sterreichischen Ausgrabungen am Kordlar-Tepe in Aserbaidschan.
MAGW 106: 6582.
Levine, L.
1987 The Iron Age. In Hole, F. (ed.) The Archaeology of Western Iran: 229250.
Washington, DC.
Lippert, A.
1976 Vorbericht der sterreichischen Ausgrabungen am Kordlar-Tepe in Per-
sisch-Aserbaidschan: Kampagne 1974. MAGW 106: 83111.
1979 Die sterreichischen Ausgrabungen am Kordlar-Tepe in Persisch-Westaser-
baidschan (19711978). AMIran 12: 103137.
Medvedskaya, T.
1977 On the Iranian association of the grey ceramics of the early Iron Age.
Vestnik Drevnei Istorii: 93104. (In Russian).
1982 Iran: Iron Age I. Oxford (BAR International Series 126).
Meldgeard, J., Mortensen, P. and Thrane, H.

1964 Excavations at Tepe Guran, Luristan. Preliminary report of the Danish
Archaeological Expedition to Iran 1963. ActaArch 34: 97133.
Muscarella O.W.
1968 Excavations at Dinkha Tepe, 1966. BMMA. November 187196.
1974 The Iron Age at Dinkha Tepe. MMAJ 9: 3590.
1988 The background to the Luristan bronzes. In Curtis, J, (ed.) Bronzeworking
centres of Western Asia c. 1000539BC: 3344. London.
1989 Warfare at Hasanlu in the late-9th century BC. Expedition 31.2/3: 2436.
Stein, A.
1940 Old routes of Western Iran. London.
Stein, D.
1984 Khabur ware and Nuzi ware: their origin, relationship and significance. Assur
4.1. Malibu.
Young, T.C. Jr.

1965 A comparative ceramic chronology for Western Iran, 1500500BC. Iran 2:
5385.
1967 The Iranian migration into the Zagros. Iran 5: 1134.
1985 Early Iron Age revisited: preliminary suggestions for the re-analysis of old
constructs. In De lIndus aux Balkans, Recueil Jean Deshayes: 361378. Paris.
Figure 1. Map of Urmia area. Courtesy of the Penn Museum.

Figure 2. Hasanlu Period VI: nos. 3, 6, 1114; all others are Period V
(scale 2:5). Courtesy of the Penn Museum, image #73867.
Figure 3. Jars. Dinkha Period IV (scale 1:3).

Courtesy of the Penn Museum, image #92862.
Figure 4.1. Dinkha Period III (Iron Age I) polychrome vessels.

Figure 4.2. Dinkha III, Iron Age I burials.

Figure 5. Geoy Tepe, tomb K vessels, Iron

Age I. Courtesy of the Penn Museum.
Pl. 1.1. Dinkha polychrome sherds, Period IV.

Pl. 1.2. Iron Age I diagnostic shapes (Dinkha III).

Pl. 2.1. Iron Age II diagnostic shapes (Hasanlu

IV). Courtesy of the Penn Museum, image #139910.
Pl. 2.2. Pottery from Hajji Firuz burial, Iron

Age I. Courtesy of the Penn Museum.
Pl. 3. Tomb 7B B2, Dinkha, Iron Age I.

chapter fourteen
JIROFT AND JIROFT-ARATTA:

A REVIEW ARTICLE OF YOUSEF MADJIDZADEH,
JIROFT: THE EARLIEST ORIENTAL CIVILIZATION *
Dedicated to my friend and colleague

T. Cuyler Young, Jr.
Rudyard Kiplings poem (from The Elephants Child) has been with me for
yearsand I have previously cited it.1 In researching the Jiroft story it resur-
faced in my mind because the epistemological questions it poses are per-
tinent to the following discussion, and thus I (appropriately) cite it again
here:
I keep six honest serving-men
(They taught me all I knew);
Their names are What and Why and When
And How and Where and Who.
I send them over land and sea,
I send them east and west;
For they are hungry men,
But different folk have different views
Indeed; and it is the different views of the issues of asking or not asking,
and not answering, the questions raised by the honest serving men that
generates this review.
1. Introduction
Archaeologists have learned about a collection of remarkable unexcavated

artifacts that were confiscated in Iran and accepted by many to have derived
from the Jiroft area in south central Iran. They became aware of these
* This article originally appeared as Jiroft and JiroftAratta: A Review Article of
Yousef Madjidzadeh, Jiroft: The Earliest Oriental Civilization Bulletin of the Asia Institute 15
(2001): 173198.
1 See Bronze and Iron (New York: Metropolitan Museum of Art, 1988), p. 9.
486 chapter fourteen
objects from a number of sources: online Iranian news services beginning in

2001, Iranian Archaeological News Bulletin (IANB), Islamic Republic News
Agency (IRNA), and the Persian Morning Daily; other online news services;
reports in two French archaeological magazines published in April (fig. 1)
and October 2003 (2003a, 2003b), the latter fully dedicated to Jiroft with
many photographs and articles by a number of scholars; three articles by
A. Lawler (2003, 2004a, b); and in the United States public lectures have also
commenced. In May of 2003 I began email discussions on Jiroft with some
scholars in Iran and the U.S.
Madjidzadehs book was published in Tehran in June 2003, after the
French magazine report in 2003a, but before 2003b (I acquired it in April
2004).2 Mentioned in some of the sources above but not here are brief dis-
cussions about excavations recently begun by Madjidzadeh in the Jiroft area
(see 2003b, 65ff.) at Konar Sandal B; unnamed in Lawler 2004a, 40 ff. Archae-
ological activity here, however, is just at the beginning stage of anticipated
long-range excavation. Madjidzadeh 2003 is a casual, brief discussion in
eight pages in English (pp. 512, with a French translation, pp. 1319) about
the confiscated material, accompanied by 163 plates of photographs (pp. 11
174) and a catalogue giving shape, material, and measurements of a selection
of the confiscated corpus (pp. 175218; see, for example, additional confis-
cated vessels in 2003b, 85); its paper jacket is superbly designed, most attrac-
tive. My evaluation and comments of the corpus follow in the text below,
but it is pertinent to note up front that what we have is a hastily assem-
bled catalogue-picture book, not an archaeological report. Vouchsafed are
only a few brief, inadequate, descriptions of some of the published objects;
and less than a page dedicated to mentioningbut not evaluatinghow
the objects came into existence in the first place (other than to note that
confiscation occurred). Utterly missing is awareness of the total absence of
an archaeological background associated with the confiscated corpus and
a perception of its limited value as data for archaeological and cultural
information. Unspoken, because assumed in the text, is that all the confis-
cated artifacts represent the ancient history of Jiroft (area)a datum pre-
sented also in the scholarship offered in 2003a and 2003b. Indeed, although
many of the objects have been published elsewhere (2003b) the photograph
2 On the book spine and its cover the authors name is spelt Madjidzadeh, but in the three
pre-text publisher pages it is spelt Majidzadeh. I use the former spelling because libraries
will use it. Fred Hiebert kindly sent me copies of photographs previous to my acquiring the
volumeas a gift from from Ali Vahdati in Tehran.
jiroft and jiroft-aratta 487
Figure 1. Archologia, April 2003, cover.
record here in one venue is valuable and important for continuous exami-
nation and research.
Up to the time of completing this review (August 2004) no excavated
artifacts have been reported from Madjidzadehs excavation (it is not a
cemetery site) that relate or can be compared to the dissimilar, unexcavated
objects published in his Catalogue. Therefore, only these latter objects per
se are capable of being analyzedand which objects, I argue, must in
archaeological discourse be cited and labeled Jiroft, not Jiroft, artifacts.
An evaluation of information and assertions about Jiroft artifacts and the
nature of their ancient culture explained in the 2003 Catalogue and the
other published reports forces one to confront (again!) basic archaeolog-
ical methodology, here concerning cultural and historical interpretations
and evaluations involving unexcavated artifacts known solely from mul-
tiple confiscations, and with no records preserved of these acquisitions.
Revealed is this, at present there is no knowledge of Jiroft (i.e., from excava-
tions in the Jiroft area), and very little available of Jiroft except a number
of objects. Between the two conceptions there is a profound epistemolog-
ical and archaeological difference (but one would never know this from
the Jiroft literature: alleged archaeological attempts at conclusion-forming
procedures have altered very little in the last fifty years).
2. History of Jiroft
Date of the Plunder

Most reports agree that flooding in the Jiroft area revealed ancient tombs
leading to plunder at a number of cemeteries, but there is no consensus con-
cerning precisely when the plundering first began: Madjidzadeh in 2003a,
37, gives no date except that of an incident in early 2002; in 2003b, 23,
a specific date is mentioned for initial plundering Au dbut de l anne
2001 a time repeated in the Catalogue (2003, 6) as February 2001; in IRNA
online 7/29/03 the date is mid 2001; in Lawler 2004a, 46, the date is sim-
ply 2001, but in Lawler 2004b, 50, it was in 2000. Inasmuch as archaeolo-
gists and others eventually viewed the plundered areas, and photographs
of destroyed tombs do exist (viz. Lawler 2004a, 44), it is a fact (one of
the few in the whole story; see below) that mass plundering occurred at
a number of sites in the southern Kerman province, in the area of Jiroft.
A photograph of plundering in action is even publishedbut, unfortu-
nately, without attribution or date (Madjidzadeh in 2003b, 22; Lawler 2004b,
50); and an Iranian scholar informed me in June 2003 that he had vis-
ited a cemetery site (but he did not say when) in the Jiroft area while
still a couple of hundred of people were still busy with illegal excavation
In IRNA online 7/29/03, Abdolali Hessam Arefi stated that plundering
was still in progress. For what purports to be a diachronic history of the
initial plundering, the continuing plundering, and the difficulties of local
and national police and archaeological authorities over two years, we have
the article by Hamid-Reza Husseini in the Persian Morning Daily online,
9/8/02.
As if the same script writer were involved, the history of the discovery and
confiscation of the Jiroft material uncannily parallels in all formal details,
including local professional inaction, the discovery and partial confiscation
of the plundered Kalmakarra cave material in western Iran: see Henkelman
2003, 214ff. To his credit, Henkelman also distinguishes between Kalmakarra
and Kalmakarra objects; see below.
In the Bonhams antiquity sale of 9/22/88, at least nos. 171, and 172, a
sculpted snake vessel and another vessel with cut-out rosette decoration
are neatly paralleled in the Madjidzadeh Catalogue: pp. 109 and 115. This
may indicate that plundering in the Jiroft area began much earlier than
reported, at least sporadically, or that these objects have a wide geographical
distribution.
Dates and Venues of Confiscations

According to Madjidzadeh (in 2003a, 37), confiscations occurred in Febru-
ary 2002; in 2003b (p. 25) the date February 2002 is also mentioned as the
time the police forces came into the area to protect it, and he gives the
names of four towns and cities where confiscations occurred: Bardsir, Jiroft,
Bandar Abbas, and Tehran; in the Catalogue (2003, 6), however, the date
February 2001 is given as the time police arrested several smugglers and
confiscated artifacts in Jiroft and Bardsir, and other cities. A 2001 date is
supported by information reported in several online reports, although only
one confiscation locus was mentioned, Jiroft. IANB 11/22/01 reports that 60
objects had been seized, in 12/13/01 that 120 objects had been seized; later, in
IRNA 7/29/03, 50 objects are reported to have been confiscated in the pre-
vious three months. An Iranian colleague (Ali Vahdati) informs me that 33
objects were also recently (2004?) confiscated at Qum, south of Teheran. Le
Monde online, 10/2/03, reports that in 2002 Madjidzadeh saw confiscated
objects in the Kerman prison (their confiscation date is not mentioned),
which suggests that sometime in 2002at least a year after the time of
the initial plunderingarchaeological authorities first became involved.
Le Bien Publique online (also 10/2/03) mentioned that customs agents had
seized two trucks (camions) filled with antiquities; the date and locus
of this seizure were not mentioned, but the report stated that this seizure
led to the discovery of le site [sic] de Jiroft. Concerning a one-site iden-
tification, note that almost every archaeologist who discussed the Jiroft
artifacts publicly (lectures) or in emails consistently mentioned them as if

only one site were involved.
From the above information it seems then that just as with the time
period of the plundering, the initial time of the confiscations is not remem-
bered. Confiscations certainly began in 2001 and continued into 2002 and
2003and later. But no information has been presented that indicates if
the confiscations at the four loci mentioned occurred within days, weeks,
or months of each other. And it seems that it was not until sometime in
2002 that archaeological authorities became involved in investigating the
plunderingwhich had been in progress for a year or more.
Who Witnessed and Recorded the Confiscations?

Nothing in the published sources vouchsafes information or interest con-
cerning: how the multiple confiscations were accomplished (we have only
one mention of trucks, but no geography)i.e., whether by means of fortu-
itous or accidental information, via informants, voluntary surrender, etc.;
which specific objects were confiscated at which of the towns and cities
mentioned above; and what were the specific venues of the confiscations
(aside from trucks). I raised these matters with colleagues monitoring the
issues both in Iran and the U.S., stressing their importance and asking for
answers. A scholar/administrator in Iran responded that it was impossi-
ble to answer the questions, that the confiscations were accomplished by
different people belonging to various organizations. No records of the con-
fiscation process or the nature of the venues were kept. An archaeologist
involved in Jiroft issues wrote to me (email) vigorously you are not being
objective or even rational The information you demand will NOT be forth-
coming! villagers looted a [sic] site, local [sic] authorities confiscated the
collection it follows that archaeological decency demands that all the
confiscated and published Jiroft artifacts be accepted as having been plun-
dered from cemeteries near Jiroft: end of discussion.
The lack of basic data specifically indicates that no information exists
that could enable archaeologists to comprehend what kinds of investiga-
tions and judgments were utilized by authorities in Teheran and Bandar
Abbas (separated by over 1,000km) that compelled the assignment of all the
material confiscated as derived from the far-away Jiroft area. No records
seem to exist that inform which specific objects were confiscated in the
Jiroft area, i.e., adjacent to the plundered areas, and which in each of the
two (three with Qum) distant citiesinformation that would reveal the cor-
pus modern cultural contexts, as well as allow analysis concerning whether
objects or clusters of one particular style were confiscated in only one or

in several of the loci. Confiscated objects are housed in a museum in the
town of Jiroft (see Lawler 2004a, 48), but were they locally confiscated
or were they transported there from Bandar Abbas or Teheran? In this
context consider the following contradictory informationa leitmotiv of
the problems discussed here: I was informed by a colleague in Teheran that
there are no Jiroft objects housed in the Iran Bastan Museum and that
all are curated in two venues in Jiroft and in the Kerman Cultural Heritage
museum. An archaeologist colleague, however, informs me that while in
Tehran he was told there were at least three vessels in the Iran Bastan
Museum, but he did not see them.
No archaeologist knows the venues where the confiscations occurred
at each locusi.e., whether in private homes, dealers shops, bazaars, etc.
Indeed, an object recovered in a dealers shop in Teheran could have had
(but not necessarily, see below) a different recent history than one recov-
ered in the Jiroft area. Lacking such background provenance information
confuses judgments regarding possible archaeological provenience and pos-
sible authenticity of a confiscated Jiroft artifact.
To date, there exists in print no empirical evidence or information about a
single Jiroft artifact, not even its confiscation history. This remains a crucial
datum that has to be addressed up front in scholarly research on Jiroft and
Jiroft. Hence, not a single one of the said to have come from Jiroft artifacts
can archaeologically [sic] be identified as in fact plundered solely in the Jiroft
region of Irannot a minor issue here. This should be the starting point
of any Jiroft/Jiroft discussion, but no scholar, including Madjidzadeh, has
seen fit to confront it.
Quantity Plundered and Confiscated

Madjidzadeh (in 2003a, 37) states that almost 1,000 objects were plundered,
in 2003b (p. 25) that local authorities claim thousands of artifacts were
plundered in the area, and later (in Lawler 2004a, 46) We guess that 100,000
objects were looted. IRNA online 7/29/03 claimed hundreds of thousands
of artifacts had been plundered. A fair statement on this matter is that no
one knows how many objects have been plundered.
Speaking to the numbers of objects confiscated, Madjidzadeh (2003, 6)
gave the figure as approximately 500 items, 300 of which were vessels; later
(in 2003b, 25) he informed us that more than 2,000 objects had been con-
fiscated. In his Catalogue (2003, 7) he is publishing a large number of the
objects recovered from the illicit excavations in the region of Jiroft, which
comes to about 260 objects. Lawler (2003, 974) records that many hundreds
of vessels were confiscatedhow many were plain (a number of which
were confiscated, see Madjidzadeh 2003, 159, 163), and how many decorated
with motifs, is not revealed. We await an inventory of the confiscated objects.
Named Plundered Sites

In Iranian Archaeological News Agency 11/22/01 and 12/13/01 online, objects
were said to have been unearthed in the old city of Jiroft Madjidzadeh
2003a, 37, also named Jiroft alone as the plundered site. And other schol-
arsviz. J. Perrot and S. Cleuziou (in 2003b) also refer to only one site, Jiroft,
as have scholars in the U.S. and Tehran in email messages, who mention the
site, the cemetery, one cemeteryat Jiroft. Notwithstanding the one-site
issue, Madjidzadeh (in 2003, 6 and 2003b, 25) names as the most important
of the plundered sites, five situated from 29 to 53 km to the south of Jiroft
(Jiroft itself is not mentioned here as a plundered site).
Excavated Tombs
Tombs have been excavated by an archaeologist (H. Choubak) at Riganbar
(one of the five named plundered sites). Although Madjidzadeh publishes a
photograph of one with its burial goods intact (in 2003b, 25), nothing of its
contents is mentioned. Ali Vahdati informs me from Teheran that this tomb
was the only Bronze Age tomb excavated, and that the others are Islamic,
and he confirmed my suggestion that no decorated vessels were recovered
here. Worth noting is that Madjidzadeh is engaged in an excavation strategy
of digging only a settlement site, not searching or excavating the plundered
cemeteries (Lawler 2004a, 46). Such unilateral action ignores the precious
work and model of the great Belgian archaeologist Louis Vanden Berghe,
who spent fifteen years surveying cemeteries in Luristan that had been
plundered for decadesand thereby recovered hundreds of intact burials.
His model should be followed, not ignoredeven one tomb excavated with
Catalogue material would be a very significant archaeological discovery for
Jiroft!
From information gathered from local plunderers regarding the number
of objects recovered from the plundered tombs Madjidzadeh (2003,6) says,
in many cases a single grave contained up to sixty objects. And from Lawler
(2004a, 46) we are further informed from the same local sources that each
grave contained at least one stone vessel; the largest one contained 30. I
think it not unfair to say that these inventory records are hearsay, and may
not be cited as historical reality.
Jiroft Style
There are several different styles, depictions, and a variety of iconographies
represented in the confiscated corpusespecially articulated on bowls,
pedestal footed vessels, beakers, and weights. Many of these styles and
iconographies have not hitherto been encountered (strange and different,
having many entirely new items as one archaeologistwho defends the
corpus integrityaccurately reported to me). Perrot (in 2003b, 111) uses the
phrase style de Jiroft, but styles in the plural is a more accurate term, given
that for many entirely new items parallels do not exist, except in Jiroft.
Madjidzadeh (in 2003b, 26) reported that clandestine diggers reported to
him that relief decorated vessels were recovered, a claim again, which (here
parti pris) has no archaeological value. And no one to my knowledge has
reported finding a decorated vessel in surveys.
One archaeologist believed at one time (email message to me) that there
is a virtual absence of the classic intercultural style but later shifted
entirely, claiming that the vast [sic] majority are indeed of the IS!
Cleuziou (in 2003b, 116C 122) notes differences between Jiroft and Tarut
artifacts, but stresses stylistic parallels (viz. entwined snakes); and Mad-
jidzadeh (2003, 7) places all the chlorite objects in the srie anciennes, i.e.,
early Intercultural Style (IS).3 A good number of the confiscated objects are
canonical members of the IS corpus, or readily relate to them: see Mad-
jidzadeh 2003 for imbricate, whorl, spiral, palm tree, hut, and animal and
scorpion patterns: pp. 44,6775, 110111, 117118, 125, 127129, 142; also the
vessels illustrated in Pittman (in 2003b, middle and bottom of p. 85; also
Cleuziou 2003b, 117, fig. 4, and 122). For convenient excavated examples for
these and other motifs in the IS corpus see of course Burkholder 1971, Kohl
1975 and 2001, Zarins 1978, and Lamberg-Karlovsky 1988.
According to Madjidzadeh the corpus reveals a high quality of work-
manship, (2003a, 37, also in 2003, 8), but he is also aware (2003, 10) that
it is not always equal, that (here no doubt reflecting the multiple styles
and iconographies) the corpus was produced in different workshops, and
by different stone-cutters, having different levels of skill and talent. But, in
comparison with the Mesopotamian reliefs in stone, they appear in general
to be of superior skill, talent and capability. He does not explain or develop
any of these important cultural and aesthetic issues further, give specific
examples, or discuss whether the different workshops recognized could be
3 For a defense of the term Intercultural Style, see Kohl 2001, 209, 215f.
correlated with different confiscation loci. But no one can dispute his obser-
vations on the skill and workshop issues.
(In a discussion with Philip Kohl an idea occurred: stone analyses should
be undertaken of both Yahya and Jiroft stone material to determine
sourceswe know there was one near Yahya. I see no evidence that one
can claim that Yahya craftsmen manufactured the latters stone artifacts.)
Chronology
Since most archaeologists involved in Jiroft accept that the published cor-
pus derived from (somewhere around) ancient Jiroft, chronological ranges
may be estimated and have indeed been proposed. Madjidzadeh (2003a,
37, 44; 2003, pp. 7, 12) asserted an early date for chlorite vessel production,
3000bc, or late 4th to first half of the 3rd millennium bc; this chronology
is accepted by Perrotil y a 5000 ans, 31002900bc (in 2003b, 97, Ill).
Cleuziou (ibid., 116) raises the question whether this early chronology is cor-
rect or a later date, Early Dynastic IIIII, even into the Akkadian period, is
better, but seems to favor the later dating; Pittman (ibid., 81, 87) accepts a
general 3rd millennium date; Lawler (2004b, 50) reports (from an anony-
mous source) an around 2500 bc date. Recently P. Amiet (2002/2004, 95 f.)
rejected Madjidzadehs early dating, arguing for the late 3rd millennium.
He is correct; and one would expect that the genuine IS period artifacts in
the corpus be invoked for chronological determinations, comparing them to
excavated material from Mesopotamia and Yahya, and that recovered from
Tarut.
Madjidzadeh, however, knows (and therefore need not explain) that the
Mesopotamian artifacts post-date those from Jiroft and Yahya. But dis-
interested analysis indicates that his and Perrots beginning and flourit
chronology of Jiroft is fundamentally wrong, too high by more than a half
millennium. Concerning at least the IS material, the second half of the 3rd
millennium bc, which includes the late Early Dynastic and at least part of
the Akkadian period, is an accurate general chronology. Lamberg-Karlovsky
(1988, 54) sees IS material continuing into post-Akkadian times, as does
Kohl (1975, 30; idem 2001, 215, 220f., 222, 224, 226 f.), where it is claimed
that at Yahya IS material was made late in the style, in Akkadian or even
post-Akkadian times. One of the issues here is recovery of IS objects in post-
Early Dynastic, Akkadian contexts: see Martin in Aruz and Wallenfels 2003,
no. 233, an IS feline-snake combat scene in Berlin bearing an inscription
signed by Rimush, Lugal Kish. This object, albeit not excavated, is quite
important for establishing the chronological range for these scenes.
What must be faced unblinkingly here is that the early unanchored chro-
nology is generated not by archaeological reasoning, but by a priori tenden-
tious, self-serving conclusion formations about Jiroft and civilization.
Jiroft and Civilization

Quite soon after the appearance of Jiroft artifacts the pitch of interpre-
tative language was set very high. Madjidzadehs Iranian to Mesopotamian
relative evaluation is but a minor ingredient in the Jiroft brew, but in the
same publication, beginning on the first page (2003, 5 f.), he raised the level
of cultural evaluation higher. The hyperbole of the Catalogues title of course
warns us to anticipate the instructions: that the confiscated objects compel
us to review our current ideas about the origins of the Mesopotamian, and in
particular the early Sumerian Civilization, and that the objects clearly sug-
gest that a considerable part of the Sumerian art may have originated in the
southeastern Iran, in the region of Kerman, and (ibid., 12; also in Lawler
2004a, 43f., 48) that Jiroft (the Kerman area) is to be recognized as the Land
of Aratta, that mysterious civilization Not unexpectedly, no specific
objects or parallels are presented (compare Cleuzious 2003b, 116, 117, and
122, more rational approach regarding specific Mesopotamian and Jiroft
parallels and differences). There is more. Madjidzadeh (2003, 11), mention-
ing but not supplying recognizable iconographic parallels, informs us that
the Sumerian Etana myth originated in Iran and traveled west. Hyperboles
easily spawn others. C.C. Lamberg-Karlovsky in Le Monde (online 10/2/03)
asserts that Jiroft calls into question our fundamental [sic!] concept of the
origins [sic] of the age of the Middle East; and in Lawler 2003, 973, shares his
belief that From now on, we must speak of before and after Jiroft, a conceit
echoed by Perrot (in 2003b, 111). This is heavy stuff indeed.
This rhetoric is presented to the archaeological and public communities
as a given, a perceived manifest fundamental truthnotwithstanding it
is not anchored in empirical archaeological arguments or chronological
analyses, and it does not consider that the results of site excavations lie in
the future. For, regardless of what the significance of the Jiroft areas culture
and chronology may be, manifestly without the benefit of excavations there
can be no justification for the present rhetoric and hyperbole broadcast.
Accepting as archaeologically reasonable that the genuine artifacts con-
fiscated in the Jiroft area most probably were plundered there, H. Pittmans
nuanced assessment is closer to a meaningful appraisal. She compares the
Jiroft culture formally to that of Mesopotamia and the Indus, but notes that
it is smaller in scale and less complex (in Lawler 2003, 974), and (in Lawler
2004a, 42) rejects Madjidzadehs civilization claims. I too have argued that
the ecstatic claims under review do not reflect reality, that even if all the
Jiroft material were ancient artifacts, it is not world-shaking (in Lawler
2004a, 48with my original phrase restored).
3. Ancient and Modern Jiroft
Are all Jiroft objects manifestly of ancient manufacture? When I first en-
countered the Jiroft material in a lecture by H. Pittman in April 2003 (I
heard another in March 2004), aside from typical IS present, I was surprised
by the variety of styles, iconographies, and forms other than what is known
either in Mesopotamia or Iran, including the Kerman area. Locally, Yahya
shares the typical geometric IS non-representative motifsthe snake-lion
and snake-snake combat, whorl, and scorpion representations (see Kohl
1975 for patterns with locus map, p. 24, and p. 26, nos. 2, 3, 4; Kohl 2001, 222,
226, figs. 9.8, 9.13; Lamberg-Karlovsky 1988, 55 ff., figs. 1, 3, 4, pls. IVX; omit
from both articles references to Azerbaijan and Palmyra as excavated sites
yielding IS objects). My sense, however, was that a number of objects could
not automatically be accepted as ancient productions merely because they
had been confiscated in Iran (of which action little is known). These views
were reinforced after further reading, but I readily admit that with ongoing
study over time I have changed my opinions several times regarding the ages
of specific objects.4
Judging from the publications and personal discussions, raising the issue
of possible forgeries in the Jiroft corpus is not a popular opinion. The fervor
aroused in some who accept the whole corpus as genuine may be under-
stood from email comments by an archaeologist with whom I exchanged
views regarding ancient or modern. Initially believing many objects to be
impossible, my correspondent noted that this position was soon reversed:
4 I experienced similar views in the past regarding then strange iconographys surfacing
on the antiquities market that related them to IS objects (viz. snake-lion confrontations),
and raised the issue of whether they are ancient or not (Muscarella 2000a, 171). In some
instances an abeyant cautious view was recommended ibid., 171, nos. 16, 18af. For the
record, of the complex no. 16, I am now more at ease; I also believe in the antiquity of
no. 18-eprimarily because of the execution of the guilloche/whorls, for no. 18-c, p. 488, the
iconography is not an issue, and aside from the snakes ears and body line, cannot condemn it;
pp. 169, 486, no. 7, still puzzles me but I cannot condemn it outright. Nota bene, that changing
ones mind over time is a correct and necessary activity, it results from what I call the gift of
the bazaar, the chaos caused by the antiquities market.
I accept this collection as authentic and What is in the catalogue [Mad-

jidzadeh 2003] is genuine, the corpus in toto is embraced as productions
of the 3rd millennium bc. My more hesitant opinion on this matter was
dismissed as based on personal authority absurd, as a voice crying in
the wilderness, indeed, as an opinion that matters little (in the archaeo-
logical communitythe ultimate dismissal!). Added was this psychological
insight: Your emotions dictate your perceived reality [your] objectivity
has flown the coop You have made up your mind This is your desire.
To further differentiate my views from reality, to objectively (but here not
the flown from the coop variety) document the authenticity of all the Jiroft
artifacts, a graduate student in Iran who believes that all the Jiroft artifacts
are genuine was invoked as an authority; another authority invoked was
an archaeologist who it was alleged accepts their authenticity without hes-
itation.
There are a few references in print where Jiroft forgeries are mentioned.
Pittman (in Lawler 2003, 974) accepts that the corpus derives from graves
but doesnt know if [fakes] were addedto the corpus subsequent to plun-
dering activities. In this same venue I stated a more sanguine view about
whether the full corpus derived from the plundered graves, and suggested
that we have to start at square one, that is, examine each object on its own
terms (see also Lawler 2004b, 49, 50). Amiet (2002, 96) indicts two objects,
one as doubtful (Madjidzadeh 2003, 106) and one as an obvious forgery
(ibid., 147). And in the Art News January 2004, 9, an unnamed London dealer
is quoted that he is worried about the growing number of fake Jiroft vases
now circulating on the market. Gasp! What fakes could this man possibly be
talking about? What does this dealer know that authorities do not know?
A concern for forgeries is relevant here given that some time had elapsed
between the time of the initial plunder and confiscation (which I am in-
formed by Ali Vahdati is still occurring), at different and distant loci, and
times. Hence, a long recognized and enduring modern cultural activity
demands consideration: forgers begin work immediately after significant
archaeological or plundered or confiscated finds become known, viz. the
forgeries that surfaced along with genuine material said to come from (by
dealers, curators, and collectorsand yes, also by archaeologists) Ziwiye,
Luristan, Hacilar, Marlik, etc. The most recent manifestation of this activity
is the existence of forgeries associated with the Kalmakarra cave objects that
were also plundered from Iran (Muscarella 2000; see also Henkelman 2003,
214f. and n. 120). Henkelman reports that among the Kalmakarra objects
confiscated in Iran some seemed suspicious, and these are in addition to
those objects smuggled abroad that he correctly identifies as forgeries.
To this unending list we now add Jiroft, another modern construc-

tion of archaeological scholarship. That forgeries could have been made in
Iran, in the Jiroft region itself or elsewhere, is viable, possible, most cer-
tainly not impossible. This would explain how Jiroft forgeries could have
been collected at any of the confiscation loci in Iran together with gen-
uine ancient materialthe confiscation occurring before smuggling abroad
commenced. Jiroft objects surfacing abroad could thus include both gen-
uine and forgeries smuggled out together to present a found together
corpus(a topos beloved by the collecting world of dealers, museum cura-
tors, and private collectors). No dealer or smuggler would be stupid enough
to smuggle out forgeries and genuine objects in separate shipments. (After
this paper was essentially completed, Ali Vahdati informed me [June 2004]
that a colleague working in the Jiroft area told him that the police caught
a local smuggler of stone vessels. The prisoner vigorously denied plunder-
ing the vessels from a siteno, he was not a thief, he was an artist! From
the plunderers he purchased plain vessels that were worth very little money,
decorated them with motifs that were copied from originals, and sold them
at a good price. Vahdatis source said the copies were quite good, one could
hardly distinguish [them] from the original ones. I am not surprised, but
until one sees the alleged forgeries and originals together I refrain from
comment.)
Starting from square one means that all the confiscated, aka unexca-
vated, objects receive a close reading, a stylistic examination, questions
asked why are they ancient, and viable answers attempted: put another way,
go back to Kiplings honest serving mens questionsI cannot state it better.
All the more so when one focuses on the Jiroft style/iconography/work-
manship issues, which are barely discussed in the literature: because inas-
much as all Jiroft objects are ancient, why waste time? But judging disinter-
estedly from known excavated artifacts (aka archaeological data), a number
of the confiscated carved and unique representations of humans, animals,
and flora appear to be crudely portrayed, stylized in execution, unskill-
fully and non-uniformly (even within one scene) executed, especially hands,
beards, eyes, feet, noses, etc. Stylistic analyses joined to its modern acquisi-
tion history collectively suggest that it is impossible to declare that every
object in the Jiroft published corpus is necessarily an ancient artifact
although a good number are indeed ancient, and are listed below (see a,
below).
I cannot assert that any given object in the corpus is absolutely a forgery:
not because I doubt what my eyes and knowledge tell me, or that I underes-
timate the skill and knowledge of forgers (the best are guaranteed good pay
and lifetime jobs), but because I cannot claim to know, to perceive all the
possible scenarios in their ancient/modern histories. Hesitancy is formal, an
attempt to keep all options open, one of which is the possible presence of
forgeries. An archaeologist experienced in IS and Iranian scholarship shared
this view with me: I would argue that it would be almost impossible to
decide whether most [Jiroft] objects were genuine or fakes. Which neatly
defines the problem I am articulating here: absent a Jiroft against which
to compare Jiroft, one is compelled to focus on problematic orphaned
objects. If we do not attempt to know which Jiroft objects are ancient and
which modern, how can we begin to discuss Jiroft?
A number of Jiroft objects stand out from the IS corpus, suggesting (at
least) that archaeologists not automatically accept and introduce them into
discourse on ancient artifacts; these are listed below as what at best are
called probable forgeries (below, b).
In addition, my eyes recognize a number of complex and ambitiously
made pieces with which I continuously wrestle (see below, c). They may
not be ancientbut I have changed my mind more than once with regard to
the age of several of them. Given this unclear viewa mon avis of courseI
propose that they be kept in abeyance, subject to physical and continuous
stylistic analysis. And cited within parentheses, with a caveat. The objects
themselves are not numbered and can be identified only by page references
in Madjidzadehs Catalogue (which I employ here). I have not autopsied a
single object from the Jiroft corpus, but note that those who accepted the
corpus objects originally did so also from photographs.
In the discussions I list some examples of possible Jiroft objects that
have surfaced outside of Iran and appearing for sale in auction and dealer
catalogues. If I am correct about attribution they should be considered as
additions to the corpus. A number of objects offered for saleboth ancient
and forgeriesthat may have derived from Jiroft are not cited here, as I
remain uncertain about their source.
a. Ancient Jiroft Artifacts

The following objects listed in Madjidzadehs Catalogue are clearly ancient.
A number are members of the IS group, with canonical motifswhorls,
imbricates, huts, palm trees, guilloches, triangles, mat weaves on IS ves-
selsall executed by ancient artisans (for IS parallels see Burkholder 1971
objects she herself witnessed as deriving from Tarut; de Miroschedi 1973;
Kohl 1975; 2001; Jarins 1978; Lamberg-Karlovsky 1988). Some objects are
sophisticated works; others are relatively minor, plain and undecorated
works. The following list identifies a corpus that (wherever confiscated

and wherever actually plundered) forms the base from which one may
eventually add more identified genuine artifacts: pp. (1517?), 3738, 39, 40
41, 44, 5152, 53 (three objects), 57 (?), 6061, 6264 (?). 6566, 67, 68, 69, 70,
71, 72, 73, 74, 75, 7677, 8081, 8283, 86, 8788 (see c, below, no. 11), 89
90, 91, 9294, 97, 98, 99100, 103, 104, 107, 108, 109, 110, 111, (112113?), 115, 116,
117, 118 (fig. 6, bottom), 119, 120 (some), 121, 122 (note the plain, undecorated
pedestal base goblets here), 125, 127, 128, 129, 134, 137, 138, 140, (141? compare
p. 120, top), 142 to 146, 148165, 167172 top, 173, 174 (?). For some of these
objects see also Pittman in 2003b, 7885.
With regard to pp. 114115, six vessels decorated with rows of scorpions:
all examples are very ambitious in execution, and perhaps not what a forger
might undertake. They are crowded with scorpions, and none on each vessel
is executed in the same way with regard to claws, tails, wings, nor are these
units juxtaposed in the same manner. The conical vessel on the right of
p. 114 and the bowl on the right of 115 are the best made here, and seem
to be authentic; they relate to the scorpions carved on the weight, p. 127,
which is ancient; and also to an example excavated at Yahyanote the tail
pattern (Kohl 1975, 26, no. 3; Kohl 2001, 224, fig. 9.11; Lamberg-Karlovsky
1988, pl. VIII). Are the other vessels here also ancient? Probably yes. A
scorpion-decorated vessel was excavated at Nippur (Kohl 1979, fig. 3). And a
number of similarly decorated vessels exist in various collections; they are
not uncommon.
I suggest that a vessel in the Ishiguro collection is not ancient (Muscarella
2000a, 171, no. 18-a), but as reported in n. 3 here, a vessel bearing a complex
decoration that includes scorpions (ibid., no. 18-e) about which I was origi-
nally cautious I now realize is most probably ancient.
For the three raptor plaques, pp. 130, 131, 132, the only iconographical and
formal parallelsalmost exact, even to the use and position of inlaysare
the 8th century ad Visi- and Ostrogothic brooches, which naturally gener-
ate doubts (I anticipate articles on the proto-Sumerian Ostrogoths migrat-
ing [slowly] from their homeland, Jiroft/Aratta). The first two plaques are
quite close, but all three derive from different hands. The raptors heads and
beard positions of the first two are close to those of pp. 9294 (a masterpiece:
perhaps the same motive occurs on a fragment of a vessel from Uruk where
two snakes are attacked by two, not one as the authors state, raptors, Linden-
meyer and Martin 1993, 161 and pl. 68, no. 1102). These plaques are probably
ancient, unique, artifacts (otherwise we have a really first-rate forgers work
before us): but for the record, I shifted them from here to problem pieces,
section c: below, and back, several times. Again, the Jiroft problem.
b. Probable Forgeries
1. bowl, pp. 4546 (fig. 2), Hero mastering bulls: I suggest that the very poor
execution and inconsistencies of the bulls headseyes, noses, beards, bod-
ies, and the very insecurely and badly planned and executed wavy water
lines over the bulls (compare pp. 53, 125), the position of the leg tufts jut-
ting from the hooves themselves, and the mans head condemn this piece.
Compare all the details of the bulls physical forms, and especially the flow-
ing water lines above the heads, with those on a conical vessel recovered
from Tarut (recorded there by G. Burkholder; eventually it was sold to a col-
lector), and an unexcavated but genuine vessel in the British Museum (Aruz
in Aruz and Wallenfels 2003, nos. 226, 227); the Tarut vessel must have been
the model for the Jiroft object (see also below, no. 3).
See also the bull depicted in Bonhams 11/7/02, no. 196, a conical vessel
most likely from Jiroft. The vessel shape is the same as that of pp. 54
56 (see below, no. 3), and which P. Kohl informs me is not a classic IS
form.
2. pyxis, pp. 4748 (fig. 3): mermen (not intended) holding something in
each hand apparently feeding bullsfor which compare those represented
on pp. 53 and 125. Also, the uniform gross workmanship, style, and iconog-
raphy force us to reject this piece as an ancient artifact.
3. bowl, pp. 5456 (fig. 4): an ambitious iconographya human beardless

head projects above the rims of four pithoiperhaps meant to depict indi-
viduals bathing during the summers heat. But the crude and unskilled, as
well as inconsistent, execution of the heads and misunderstanding the tied-
up chignon, as well as their different placements on the vessels rims; the
different neck forms of the depicted vessels; the uneven body decorations;
and the snakes body decoration and closed-mouthed heads collectively sug-
gest a modern creation. Compare the bowl shape to that of pp. 4546 above,
no. 1.
The IS fragment recorded from Tarut (Zarins 1978, pl. 70, no. 49; Burk-
holder 1971, pl. VII, no. 21; Muscarella in Aruz and Wallenfels et al. 2003,
no. 224-e) is the model against which the present example must be com-
pared, detail for detail: the human head with chignon, the open-mouthed
snake behind. The differences are quite manifest. One could posit that
Jiroft was locally copied from Tarut (but when?). On the Tarut example
there is a vertical straight line below the human head, as opposed to the
curved one at Jiroft. It is not certain this line depicts a vesselbut one
cannot exclude the possibility. A stone fragment from Uruk depicts a figure
identified as a male with the same chignon form situated to the right of
an unidentifiable motifnot a snake (as Lindemeyer and Martin 1993, 128,
pl. 61, no. 690). These two examples document the IS chignon hairstyle, but
was the chignon gendered for men or women? (Lindenmeyer and Martins
alleged male parallels are not parallels.)
The Jiroft bowl has a neat motif-mate; it was purchased (of course in
good faith) by the Louvre (A. Benoit, Revue du Louvre, October 2003, 13 ff.,
fig. 1) after being offered for sale in Hotel Drouot, Sept 30, 2002, no. 212
(see also Cultural Heritage News online, 10/29/03). Depicted are two human
heads facing each other within a centrally placed vase that is framed by
snakes. The motif is precisely that of pp. 5456; both remain unique. The
Louvre human heads have feminine-appearing faces and a rear chignon,
all of which features seem to be correctly executedfar better, than that
of pp. 5456. Is the latter ancient?uncertain, not impossible; it is much
better executed than the bowl.
The Louvre curator Benoit (ibid., 14) does cite the parallel with pp. 5456,
as does Perrot (in Lawler 2003, 975), but in the latter publication A. Caubet
(Benoits senior curatorial colleague) is (transparently) indignant, dis-
mayed by the accusation that her purchased vessel derived from the Jiroft
plundereven though aware that the only known parallel derives from that
corpus. In fluent museum-speak both curators dissimulate, disingenuously
disassociating the (if genuine, then manifestly plundered) piece they pur-
chased from the vulgar plundering of Jiroft. Benoit notes a different color
of stone, which is not relevant to stone composition, and Cleuziou correctly
observes (2003b, 120) that its shape does not occur at Jiroft.5
Perhaps we will also be instructed that other Jiroft-like vessels recently
purchased (in good faith) by the Louvre most absolutely did not derive from
the Jiroft (or of course, as any curator knows, from any other) plundering
activity (see A. Benoit, Revue du Louvre, June 2003, 86 f., figs. 1, 2; Hotel
Drouot 2/26/03, no. 221); ratheras everyone in the Louvre knowsthey
were found by poor peasants when tilling their gardens.
5 If evidence exists, as Benoit and Caubet claim in the Louvre and Lawler articles, that
it was in a private European collection since 1968 (which I go on record as doubting: it was
not mentioned in the Hotel Drouot sales catalogue), it should be presented to archaeologists
(who do not work for the Louvre) for documentation examination.
4. p. 106 (fig. 5): vase with entwined snakes: ambitious but very crowded and
not good workmanship or execution of the heads and ears, and especially
the uncanonical presence together of both round and oval body markings.
Certainly it remains at least suspicious (as Amiet 2000, 96). Perhaps p. 103,
which has a similar scene, is ancient as here the snake heads are more
securely executed.
5. p. 118 (fig. 6): top two conical beakers: both decorated with a whorl
pattern that displays incompetent workmanshipa lack of understand-
ing of the pattern, and insecure execution of each whorl. Such qualities
I do not find among the ancient corpus (viz. de Miroschedi 1977, pl. II-
a; Zarins 1978, nos. 114, 308; Lamberg-Karlovsky 1988, 78, pl. X-b). P. 117
is a vessel with a superbly executed whorl pattern, one matched exactly
except in size by a vessel offered for sale by the Safani Gallery (Ancient
Form, 2004, 11). Two other vessels decorated with whorl patterns are in
the Jiroft corpus: Pittman in 2003b, 85, lower left and right. Both are
not made by the same hand; I am more comfortable with the latter
example.
Some other vessels with overall whorl patterns were offered for sale at
Christies, London, 5/15/02, no. 254, and the same venue, 5/13/03, no. 13. The
Louvre purchased a very obvious, stupid forgery of this motif (P. Amiet in
Revue du Louvre 1, 1987, 16, no. 7; Muscarella 2000a, 170, no. 9).
6. p. 126, weight (fig. 7): an ambitious work depicting raptor and snakes on
one side, floating scorpion men on the other. No such decorated weights
have been excavated, but the raptor motif is of course well known. Here
I see poor workmanship and inconsistencies and a lack of uniformity in
the execution and positioning of all details of the depicted figures, which
makes me reluctant to accept it without question. Note the scorpion-mens
thumbs, fingers and arms, ears, eye positions; the whorl constructions and
positions, the tails; also the raptors body, talons, beak, and beard (compare
also pp. 1517, 112113, nos. 3 and 15 in section c, below). For excavated
raptors flanked by snakes see the examples from Nippur and Tarut (Aruz in
Aruz and Wallenfels 2003, no. 234; Kohl 2001, 223, fig. 9.9; Zarins 1978, pl. 68,
no. 159), as well as Jiroft (Madjidzadeh 2003, 9294). For the snakes ear
forms, noses, mouths, compare the British Museum vessel (Aruz in Aruz and
Wallenfels 2003, no. 227), and again the Nippur and Jiroft vessels. See also
Muscarella 1993, 144, no. 1, fig. 5-a.
Figure 2. Catalogue, pp. 4546.

Figure 5. Catalogue, p. 106.


Plain, undecorated weights were also confiscated and included in the

Jiroft corpus; they are not published in Madjidzadehs Catalogue (but see
Islamic Republic News Agency 7/29/03, 3). Further, the Yahya team has
revealed the presence and availability of undecorated weights in their area
(Potts 2001, 115, 142f., figs. 4.4144.43). Such weights could (would) have
attracted forgers (local artists) who embellished them. For other examples
of probable weight forgeries or problem pieces with the raptor and snake
motif see:
a. a weight offered for sale by Christies, London 12/7/1994, no. 181 (Muscarella
2000a, 169, no. 5; see also n. 58). That it is not ancient is manifest by the
snakes teethin what looks actually like felines heads; the legs and talons
of the raptor are misrepresented; and the execution of the snakes incised
body oval decoration are not those of an ancient craftsman.
b. for the crude, inexperienced modern artisans weight on sale at Bonhams

4/13/2000, no. 300, and again in Bonhams 11/8/01, no. 201, no comment is
necessary if one looks at it for one second.
c. the weight decorated on both sides offered for sale in Hotel Drouot
2/27/03, no. 30, probably from Jiroft, and decorated on both sidesa Hero
with a tail mastering snakes, on the other a raptor-snake scene; in both cases
the snakes bodies continue onto the handle itself. This is a complex and
very skillfully made work, better than that of p. 126 and the Barakat example
(d, below)but the snakes body crossing over the raptors body puzzles
me, and the Hero has an oval inlay in his hair. Nevertheless, a fragment of a
weight handle from Yahya seems to preserve the body of a snake (it has oval
markings), but not in the same manner as the present example (Lamberg-
Karlovsky 1988, fig. 3-F, pl. IV). It may be genuine, but warrants more study.
d. the weight for sale in the Barakat Gallery catalogue vol. 32 is surely from
the Jiroft corpus, primarily because of the lion-raptor combat motifs. It
has very complex and amazingly ambitious decorations on both sides, lion-
snake combats and central figures. But a close reading of all the details
reveals many poor and inconsistent carvings: on the obverse, the central
double-headed monsters hands, shoulders, ears, necks, feet, and his stand-
ing on air; the flanking snakes awkward bodies and positioning, nose mark-
ings, mouths, tail, vertical body markings; and the lions crudely executed
not-uniform claw structures, mouths, feet, tail terminalsone is sculpted
as a unit of the animals back: compare Madjidzadehs pp. 8790models
for this weight? The same execution problems exist for the other side: the
lions feet, head, and eyes, and tails; the squatting man holds water flowing
from the addorsed bulls heads; his face, body, his leg lacking a foot, and his
kilt should be compared to an example from Tarut (Muscarella in Aruz and
Wallenfels 2003, no. 224-d), one in the British Museum (Aruz in Aruz and
Wallenfels 2003, no. 227, fig. 85), one in Japan (ibid., no. 235); and to section
c below, no. 2. This ambitious piece generates reservations, and at present
cannot automatically be accepted as ancient.
e. the weight offered for sale in Hotel Drouot, 2/23/02, no. 393, seems also to
be a member of the Jiroft corpus, but whether from this area or elsewhere
I suggest it is not an ancient artifact: examine the raptors body decoration
and beak, and the single animal in one talon; the mountain motif (compare
Aruz and Wallenfels 2003, no. 226), the isolated and upside-down moon
crescent (compare ibid., fig. 85). (In the same auction catalogue, again
whether from Jiroft or not, nos. 397 and 398 are also not ancient.)
f, g, h: three weights in private collections are probably forgeries; they are

listed in Muscarella 1993, 146, no. 3, p. 149, nos. 7, 10, figs. 7-b, 11, 13; idem
2000a, 169, nos. 2, 3, 4, also pp. 484, 485.
7. p. 139, small idol (fig. 8): is there any reason that allows one to assert it is
ancient? Note the mouth open (caught, in surprise?), the unique hairstyle
(Jiroft style), and the presence of scorpions as the figures arms.
8. disc, p. 172-b: a seemingly meaningless, apparently failed attempt to depict

something, a human face?
c. Problem Pieces
As noted, there may be more forgeries in Jiroft than listed above. A number
of objects presented in this section may in reality belong in section b,
aboveor indeed, perhaps in a. This category exists because I find it
difficult to react positively to unparalleled or inconsistently executed styles
and iconographies (merely because they were confiscated), and think more
stylistic and technical analysis is required; here too I have changed my mind
on several objects. What is on trial herethe jury is still outare the very
objects that in fact define the Jiroft Style.
1. pp. 1112, pedestal goblet (fig. 1, lower left): two Heroes master two upside-
down lions: I see the human head positions, their faces, eyes, ears, hair to be
frozen, not alive; equally so the legs, feet, clothing. I am uncomfortable with
the articulation of the felines rear leg positions, and they stand on their front
legs; also, the scorpion uncharacteristically lacks its sting. I do not know any
specific parallels.
2. pp. 1314, beaker (fig. 1, right): motif of a Hero with bulls, while another
is in the mountainous sky holding water or a rainbow, or a jumping rope?
All is seemingly satisfactory until we compare the mens heads, faces, the
nose mouth curves, hair, clothing, different belts, the articulation of their
shoulders, forearms and fingers, and the leg and feet of the seated Hero
with excavated male figures in the same kneeling position: Zarins 1978,
pls. 70, 72, no. 546; also Muscarella in Aruz and Wallenfels 2003, no. 224-d;
Aruz in ibid., no. 227. The stylistic and detail differences are apparent, and
probably instructive. The iconography of the lower bulls below a mountain
motif is closely paralleled to a genuine vessel recovered at Tarut (see also
above, b, no. 1), here lacking the male figure (Aruz in Aruz and Wallenfels
2003, no. 226, for mountain motif and bulls), but the execution skills are
quite differentwhen was the one modeled from the other? I also think
the star and crescent are poorly executed (compare Burkholder 1971, pl. IV,
no. 11). And contrast the bulls heads, double peaks at the top, water pattern,
and hump to p. 53 in the Catalogue.
Cleuziou (2003b, 122) accepts the authenticity of this vessel, comparing it
favorably to the Tarut vessel (above). I tend toward the negative, but remain
indecisivewhich is again an example of the Jiroft problem!
3. pp. 1517, bowl (fig. 9): Hero mastering scorpion-men; for the latter com-
pare p. 126 (no. 6 in section b, above), and no. 15, below. I find the human
and scorpion-mens heads difficult to judgebut I am more than not com-
pelled to accept it as probably ancient.
4. pp. 18 through 33, a difficult group to work with given their unexcavated
status and stylized elements: six pedestal goblets with grazing animals and
stylized trees. Some of these trees I find too stylized, with outlined leaf
borders: pp. 2123 (fig. 10), 2728 (fig. 1, upper left), 3031 (fig. 11)although
the trees of pp. 1820 (fig. 12), 2426, perhaps 3233, do seem more natural
with their isolated leaves. Compare the similar trees from Mari (Wilson in
Aruz and Wallenfels 2003, no. 231), and the ancient trees of the Catalogues
pp. 4041. Pp. 1820 has at least four separate animals, with their young,
in no regular order or position, depicted eating at trees. Pp. 2426 has two
separate levels of different animal groups; the trees are different. Pp. 32
33 seem well executed. These three vessels warrant more attention; they
should be examined and compared to the others I list here. They seem
easier to accept as ancient, more confidently executed and the motifs seem
naturalwas one or more the model for the others? Note that a number
of plain pedestal vessels were also confiscated (p. 122): how many of the
original were plain and later embellished with scenescopied from which
vessel?
There are similarly formed excavated tree representations in naturalistic
depictions from Mari (Aruz in Aruz and Wallenfels 2003, no. 231); see also
naturally executed palm trees on pp. 3730, 118, 127, 128 of the Catalogue;
also Muscarella in Aruz and Wallenfels 2003, nos. 225-a, b.
The Barakat Gallery has offered for sale a conical vessel with the same
basic animal and tree decoration; see also Hotel Drouot 9/30, 2002, no. 213.
5. pp. 3435, pyxis (fig. 13): the Christmas tree ball effectcompare a sim-
ilar motif on pp. 2426, the positions and executions of the raptorsone
trampled by a caprid, others hitting the groundand three different trees,
collectively bother me. But from what I can see of the execution, it looks
good: the work is possibly ancient.
6. p. 36 top, conical vessel, with vertical, onehorned animals. Its crudeness

bothers me, also the two methods of depicting the unicorns heads, as well
as the straight lines on the horn (a vessel of this form also with a unicorn
is in the California Museum of Ancient Art); but the execution of the trees
leaves seem correctly executed.
7. pp. 4243, pyxis: two lions each with a totally unique head and body-form
design. Were they added to a damaged ancient vessel to increase value? Two
conical vessels on pp. 3739 seem also unique, but their execution looks
fine and I am not compelled to indict them. Compare the executions of
the trees on all these vessels, also the date tree to pp. 110 and 111. A problem
piece.
8. pp. 4950, pedestal goblet, recumbent ibexes below trees: the ibexes eyes,
tails, horn tips, and the trees are poorly executed. A companion decorated
goblet made by a different craftsman is in the Barbier-Mueller Museum
(Amiet in Arts and Culture, fig. 15).
9. pp. 5859 (fig. 14): back-to-Back lions gored by bulls, a young bushy-tailed
animal and a raptor rests on the back the bulls, another is between them.
The inconsistency of the drawing and execution of the lions underbelly and
their claws, and the raptors wings creates doubts.
10. pp. 6566, conical vessel with a bull-leaping scene (fig. 15): I do not call
attention to this piece here because I doubt that this scene could appear
in central Iran, for the motif existed from the Aegean to the Indus Valley
(see Aruz in Aruz and Wallenfels 2003, 409 and fig. 100-c). But again
and of course at the very core of the Jiroft problemI am held by the
crudeness of the execution and pose of the figure standing on a bull while
holding a tree at the same time, his slightly raised right foot, and his bald,
speckled head. I find, on the other hand, that the leaping figure situated
between the bulls horns with his foot wrapped around the tree not difficult
to accept. The bull is tied to the tree, an original and intriguing scene, worthy
of further research. If the scene is anciently executednot impossible
then the fact that the humans contact with the trees has cultural signifi-
cance.
11. pp. 7890, and Pittman 2003b, 85, upper left: there are eight bowls each
with feline-snake encounters. I do not find it difficult to accept as ancient
the conical vessel on pp. 7677, but am less certain with the execution
on two other bowls: pp. 7879 (fig. 16), 8485, viz. the felines heads, their
aberrant claw constructions, the snakes heads and ears, the form of the
oval body decorations: compare these with the same forms or scenes on
excavated pieces from Yahya, Mari, Nippur, and Tarut (viz. Kohl 1975, 26,
no. 2, idem 2001, 222, fig. 9.8; Lamberg-Karlovsky 1988, 78, fig. 3-G, pl. IV;
Muscarella in Aruz and Wallenfels 2003, nos. 224-a, 224-e, 234; Aruz in ibid.,
no. 232 and fig. 87; Godarzi in ibid., no. 242-a; Zarins 1978, nos. 58, 135,
157, 542, 545; Kohl 1979, fig. 5). Compare also the superb snake and lion
combat scene on a fragment in Berlin, (Martin in Aruz and Wallenfels 2003,
no. 233) bearing an Akkadian inscription on its rearRimush Lugal Kish
(see above). Some of the excavated examples depict snake combats or snake
and lion combats; on the latter examples note especially the lions claw
construction.
The six vessels on pp. 8081, 8283, 86, 8788, 8990, and Pittman, above,
appear to be ancient (see section a, above; and compare the entwined
snake heads of pp. 8990, 9394, 99100 to the Khafajeh vessel cited by
Madjidzadeh 2003, p. 10, n. 8; see in addition Kohl 2001, 215, fig. 9.5). The
fragment on p. 91 is most probably ancient (are the claws unfinished?).
Representations of the same scene and style occur in the antiquities
market: Hotel Drouot 2/13/02 nos. 399, and 405purchased by the Barbier-
Mueller Museum (Amiet in Arts and Culture, fig. 1) but here the lion bites
into the snake, which to my knowledge is unique (?). A very bad forgery of
the scene is in a Japanese collection (Ishiguro?): The Ancient Orient Museum,
Tokyo 1978, no. 48, Muscarella 2000a, 171, no. 17.
12. pp. 9596, bowl: raptor and snakes (fig. 17); the placement of the serpents
across the raptors wings is I believe unique, but not impossible; and we
should expect to see a beard on the raptors chin, and is his beak too elon-
gated. But not insignificant, the snake has round, not oval, body patterns,
not present on excavated snakes, which feature catches our attention.
13. pp. 101102, bowl: what is the scene? Note the floating snakes heads and
an unidentifiable unita head, fire? Perhaps genuine, but to be kept in
abeyance.
14. p. 105, pedestal goblet (fig. 18): the bodies of the entwined combating
snakes seem to get lost in the entwining; other examples in the corpus are
better executedan example from Tarut (Burkholder 1971, pl. VII, no. 20).
Also how do we explain the line that divides the decoration on each snakes
body? A problem piece to be further investigated.
15. pp. 112113: bowl with double row of swimming-floating scorpion-men

(fig. 19): it is possible this is an ancient creation. It seems too ambitious
even for a sophisticated forger to makebut there is no consistency in the
execution of the heads, faces, and tails. Compare pp. 1517 and 126, b, no. 7,
and no. 3 above. A bowl in the Barbier-Mueller Museum and another offered
for sale (Amiet in Arts and Culture, fig. 11; Christies, London, 5/15/02, no. 265
have the very same swimming scorpion-man motif), but all were made by
different craftsmen.
16. p. 120, lower right, vase: the upper row of attempted guilloches is very
badly executed in form and spacing, and differ in each case, and is not
paralleled from the excavated corpus.
17. pp. 123 (fig. 20), 124: two weights in the form of openwork entwined
snakes. Both are entirely different in shape, body decoration, and sculptural
symmetry; no. 124 seems less finished. Models for the motif exist: a weight
from Soch in Uzbekistan, and an example in the Louvre (Muscarella 1993,
144, fig. 4, 149, no. 9; Kohl 2001, 227, fig. 9.14). Are both the Jiroft weights
ancient? Or is one ancient and the other a modern copy?
These two snake weights call our attention to a handle offered for sale in
Hotel Drouot 6/26/03, no. 113, a handle sculpted in open work that depicts
a male figure sculpted in the round from his kilt to his head. No feet are
depicted, but below the kilt there is an unparalleled unit of three triangles
at the front and a unit of four curved forms at the rear, so he cannot be
said to be kneeling conventionally; he masters two snakes, one of which is
connected to him by a strut. Of interest is that the males eyes seem to depict
a blind personconsciously, or unintended? Perhaps this weight doesnt
belong to the Jiroft corpus. It is a complex, unique, well-made objectbut
its uniqueness demands more investigation before secure acceptance.
18. p. 133, double-headed raptor plaque (fig. 21): aside from not first-rate
workmanship and five and six claws terminating in different lengths, it is
an unparalleled workwhy accept it unconditionally?
19. p. 135, scorpion plaque: doubts raise themselves with regard to this plaque
unparalleled elsewhere; and the excellent execution of the whole, especially
the face, confuses me. Why are its wings, body and tail forms and their
decoration different than the other scorpion plaque on p. 136 (fig. 22)? I am
not so secure with this scorpion plaque either but cannot outright condemn
either one.
20. p. 147, two odd, seemingly unfinished stone figurines, one a human, one
an animal head (?). Amiet (2002, 96) considered these to be forgeries. He
may be correct as they are quite formless, but who knows. In any event, they
can have no archaeological value.






21. p. 166, two resting felines, both said to be made from lapis lazuli. They
are close but not the same and quite simple, and it is difficult to form a
conclusion about their ages. Note the intentionally scarred right eye of one
lion, and the different front feet constructions, one open, the other closed.
The figures on p. 167, however, do not cause concern.
Conclusion
It is argued here that if Madjidzadeh and other scholars had approached

the issues while attentive to Kiplings honest serving mens questions, all
appropriate to archaeological discourse, this review might have been of a
different nature.
Acknowledgements
I want to thank Philip Kohl, Jean Evans, and Paul Collins for a close reading
of a manuscript of this paper and giving me intelligent insights and opinions.
Bibliography
2003a Au berceau de la civilisation orientale: Le

mystrieux pays dAratta? Archologia, April
2003:3645.
2003b Dossiers de Archologie, October 2003, Jiroft fabuleuse
dcouverte en Iran (complete issue on Jiroft with
articles by archaeologists), coordinated by J. Perrot
and Y. Madjidzadeh.
Arts and Cultures Arts and Cultures, no. 4, 2003 (Barbier-Mueller
Museum).
Amiet 2002 P. Amiet. Compte rendu of Madjidzadeh 2003.
Revue dassyriologie et drchologie orientale 96
[2004]:9596.
Aruz and Wallenfels J. Aruz and R. Wallenfels, eds. Art of the First Cities.
2003 Metropolitan Museum of Art. New York.
Burkholder 1971 G. Burkholder. Steatite Carvings from South Arabia.
ArtAs 33.4:306322.
De Miroschedi 1973 P. de Miroschedi. Vases et objets en steatite susiens
du Musee du Louvre. Cahiers de la Dlgation
Archologique Franaise en Iran 2:977.
Henkelman 2003 W. Henkelmann, Persians, Medes and Elamites:
Acculturation in the Neo-Elamite Period. In
Continuity of Empire: Assyria, Media, Persia, ed.
G.B. Lanfranchi et al. Padua.
Lamberg-Karlovsky 1988 C.C. Lamberg-Karlovsky. The Intercultural Style
Carved Vessels. IA 23:4595.
Kohl 1975 P. Kohl. Carved Chlorite Vessels Expedition
18.1:1831.
Kohl 1979 . The World Economy of West Asia in the
Third Millennium B.C. In South Asian Archaeology,
5585. Naples.
Kohl 2001 . Reflections on the Production of Chlorite at
Tepe Yahya: 25 Years Later. In Potts 2001, 209230.
Lawler 2003 A. Lawler. Jiroft Discovery Stuns Archaeologists.
Science, November 7, 2003:973974.
Lawler 2004a . Rocking the Cradle. Smithsonian, May
2004:4047.
Lawler 2004b . Iran Beckons. Archaeology 57.3:4651.
Lindemeyer and Martin E. Lindemeyer and L. Martin. Uruk, Kleinfunde III.
1993 Mainz am Rhein.
Madjidzadeh 2003 Y. Madjidzadeh. Jiroft: The Earliest Oriental
Civilization. Tehran.
Muscarella 1993 O.W. Muscarella. International Style Weights. BAI
7:143153.
Muscarella 2000a . The Lie Became Great. Groningen.
Muscarella 2000b . Excavated in the Bazaar: Ashurbanipals

Beaker. SOURCE 20.1:2937.
Potts 2001 D. Potts. Excavations at Tepe Yahya, Iran 19671975:
The Third Millennium. Peabody Museum, Harvard
University. Cambridge, Mass.
Zarins 1978 J. Zarins. Steatite Vessels in the Riyadh Museum.
ATLAL 2:6593.
chapter fifteen
SARGON IIS 8TH CAMPAIGN:

AN INTRODUCTION AND OVERVIEW *
Background
In 714 bc, the Assyrian king Sargon II (721705bc) documented his military
campaign east across the Zagros Mountains in western Iran, then north
into Urartian territory. His record survives on a now incomplete terracotta
tablet (37.5 by 24.5cm.) purchased by the Louvre Museum in 1910 from the
antiquities dealer J.E. Gejou. The tablet was published in a superb French
translation by F. Thureau-Dangin in 1912. Years later, a missing fragment of
the tablet that had been excavated at Assur in northern Mesopotamia by
German archaeologists was recognized in Berlin (lines 195236; nine lines of
the original tablet are still missing). This remarkable recovery and German
scholarly research ascertained that the Louvre tablet had in fact derived
from Assur (Meissner 1922). Most probably a workman had plundered the
tablet at Assur and sold it locally from where it was smuggled abroad to Paris.
The tablet was written as a report taking the form of a letter to the God Assur;
and its date, 714bc, is known from the eponymy name recorded on it. It is
one of the three known Assyrian Gottesbriefes, with the other two belonging
to Shalmaneser IV and Esarhaddon (Zaccagnini 1981: 264, note 3). Aside
from Thureau-Dangins translation, there is another by D.D. Luckenbill in
English (Ancient Records of Assyria and Babylon II, London 1989 7399),
and a German translation by W. Mayer (MDOG 115, 1983: 65132). For an
historical, ideological, and literary analysis of the text see Zaccagnini 1981.
In the 1940s, the letter engaged the attention of a few scholars involved in
the history and culture of Assyria, Iran and Urartu during the early first mil-
lennium bc (Rigg 1942; Wright 1943). Renewed interest and study began in
the 1970s and 1980s, generated both by extensive archaeological excavations
* This chapter originally appeared as Sargon IIs Eighth Campaign: An Introduction and
Overview, in The Eighth Campaign of Sargon II: Historical, Geographical, Literary, and Ideo-
logical Aspects, trans. Samad Elliyoun and Oscar White Muscarella (The Hasanlu Translation
Project, Tehern, 2012), 59.
524 chapter fifteen
and site surveys in northwestern Iran. Relevant excavations in the region

were conducted by R.H. Dyson, Jr., O.W. Muscarella, T. Cuyler Young, Jr.,
C. Burney, and W. Kleiss. The surveys were conducted mainly by Wolfram
Kleiss and Stephan Kroll, whose years of pioneering and arduous work
brought forth much information on western Iran in the ancient and Islamic
periods (regularly published in AMI from the 1960s to the 1990s). Collec-
tively, the excavations and surveys revealed what had not been previously
known, that from the 9th to the 7th centuries bc Urartian hegemony, the
Urartian State, extended beyond northeastern Anatolia and included the
western and southern shores of Lake Urmia beyond the Zagros Mountains in
northwestern Iran. In consequence of this new information, Sargons entry
into western and northwestern Iran took on a new geographical dimension.
The geographical loci and identifications of the polities and sites recorded
in Sargons letter, both non-Urartian and Urartian, and thus also Thureau-
Dangins reconstruction of his campaign routes had to be reevaluated. Sar-
gons entry into west and northwestern Iran, therefore, had to be reconsid-
ered we well. And indeed this issue became important for all future scholarly
investigations. These are primarily concerned with identifying the scores
of named polities and sites recorded in the tablet vis vis modern districts
and sites, and then attempting to trace Sargons route through western and
northwestern Iran.
Polity and City-Site Identifications
Sargons journey was long, lasting for months. Most scholars situate his
Urartian campaign, his terminal target, in the Lake Urmia region, com-
mencing either at its western or eastern shores (see Zimansky 1990:19, and
note 78); some others believe (with Thureau-Dangin) that Sargon continued
his campaign from Lake Urmia to Lake Van, the Urartian heartland. How-
ever, identifications of site and region locations presented by scholars vary,
sometimes considerably, all equally based on modern archaeological and
survey data. Scholars who have discussed these issues extensively include
Mayer, Levine, Muscarella, Salvini, Kroll, Kleiss, Burney, Zimansky, Reade,
ilingiroglu, and Liebig. Some basically agree on the identification of the
geographical locations, others strongly disagree, positing distinctly differ-
ent geographical locations for places in the text (viz. Levine 1977: 141142,
145, map; Muscarella 1986, and maps figs. 2, 3; Piotrovskii 1959: 105; Zimansky
1990: 1416: who tentatively suggest(s) his geographical attributions). Spe-
cific archaeologically known Urartian sites have been identified with their
sargon iis 8th campaign: an introduction and overview 525
ancient names, here too with different proposed identifications. I mention

only two examples, Uishe/Uaiais and Ulhu. Both sites are clearly in Urartian
territory, and identifying their precise locations is crucial for determining
the exact compass, geographical itinerary of Sargon once he was manifestly
in Urartian territory. Proposed identifications for these sites can be summa-
rized as follows:
Ulhu, Rusas royal residence (line 216) in northwestern Iran

1. Northeast of Urmia, northwest of Tabriz, Marand/Livar (following
Thureau-Dangin): L.-Haupt 1916; Burney 1972: 140, 154; ilingiroglu
19761977: 260.
2. The site of Qalatgah, southwest of Lake Urmia: Muscarella 1986; Salvini
1995: 93.
3. The Salmas plain, northwestern area of Lake Urmia, Haftavan Tepe:
Kleiss 1977: 140; but see Kleiss 19691970: 131132, fig. 3: north of Lake
Urmia.
Uishe/Uaiais
1. Ushnu area, southwest shore area of Urmia: van Loon 1975: 206207;
Zimansky 1990,16; Liebig 1991: 31, 33.
2. The site of Qaleh Ismail Aga, west ofUrmia: Pecorrella and Salvini:
1984: 2425, 46, 54; Muscarella 1986: 472; Salvini 1995: 87.
3. The site of Qalatgah: Zimansky 1985: 6667, 112, note 64; ibid. 1990:
1718; Kroll 1997: 206. According to van loon 1975: 205, the Qalatgah
inscription may mention Uise; Salvini 1984: 12, 24 doubts this, as dam-
age exists in this part of the text.
4. Sahend, east of Lake Urmia (i.e. opposite coast from Qalatgah): Kleiss
1969/70: 133.
5. Bashkale area, ca. 100km west of northern Lake Urmia: ilingiroglu
19761977; Levine 1977: 143, 147.
6. Lake Van area (Bitlis near its SW shore): Piotrovskii 1959: 105; Burney
1972: 156 (following Thureau-Dangin).
The Route of Sargon
Before his venture against Urartu in the region of Lake Urmia, Sargon had
campaigned extensively south of the lake. Scholarly interpretations regard-
ing the precise directions he undertook here, how far and where in west
ern Iran, east and north, and the approximate locations of the many
526 chapter fifteen
non-Urartian polities and districts he conquered differ considerably (see the

various discussions and route maps published). Among the many polities,
cities and districts discussed in attempts at identifications I mention but a
few. Andia and Zirkitu, are important inasmuch as Sargon states that it was
from these polities he turned towards Urartu (line 14), and thus they are
manifestly south of Urartian territory, south of Lake Urmia. Uishdish and
Subi, recorded in sequence, are also border districts, closer to Urartian terri-
tory; Subi is apparently closer, and both may have been districts situated in
or adjacent to Mannea. Sangibutu is south of Armarili, which polities were
surely south/southwest of Lake Urmia. Thus, I basically agree with Levine
1977. Armarili is a border district close to Urartu (Tashtepe, close to the lakes
southeastern corner is not excluded here). Hubushkia is a district connected
with Sargons return route to Assyria, and thus is probably west or southwest
of Lake Urmia (Reade 1994: 186, 188; 1995: 38). Recently a rock inscription of
Arghisti (785/780756bc) was recovered at Javangaleh on the eastern shore
of Lake Urmia, about 40km. north of Tashtepe (see M. Salvini in SMEA,
forthcoming). Mannea is mentioned: but this reference does not per se indi-
cate that this polity existed in the Javangaleh area, as it could refer to the
direction of an expedition and not necessarily refer to its geographical loca-
tion.
Complicating these matters is that Urartian artistic influence does occur
in the un excavated artifacts claimed by dealers and many scholars to have
derived from Ziwiye, ca. 100km. southeast of Lake Urmia (Muscarella 1977:
206207). But the only excavated Urartian artifact from this site is a seal
of Rusa I (730714bc), the protagonist of Sargon, recovered by Nosratol-
lah Motamed, the Iranian archaeologist. Furthermore, at Kull Tarike, near
Ziwiye, a tomb with Urartian fibulae was recovered (was an Urartian mer-
chant or envoy actually buried here, or were the fibulae gifted to the descen-
dant?). And an inscription of the 9th century Urartian king Ishpuini, the
father of Menua, was discovered at Traghe, near Bukan, northwest of Ziwiye.
An inscription of both was also recovered at Qalatgah, and one by Menua
was recorded at Tashtepe (Muscarella 2010: 265, 267). But these southern
occurrences probably indicate political, diplomatic interactions with
Urartus neighbors in that area, not necessarily an indication of Urartian
hegemony (see Lanfranchi 2003: 101103; Fales 2003: 140142; Muscarella
2010: 267).
With regard to determining manifest Urartian territory in the western
and southern areas adjacent to the shores of Lake Urmia, scholars agree.
But following upon this reality, the main, but not the only, disagreement
among scholars is whether Sargon entered Urartu north along the western
shore and neighborhoods of the lake before returning south, or whether he

encircled the lake, moving up its eastern shore, in manifestly non-Urartian
territory, then turned west and finally south along the lake. If the latter, then
a number of the non-Urartian polities and districts mentioned by Sargon are
to be situated to the east, not south, of Lake Urmia (viz. Zimansky 1990). A
further issue relating to these matters is that some scholars contend that Sar-
gon crossed the Zagros Mountains and reached, or even encircled, Lake Van,
then returned east by recrossing the mountains back to Lake Urmia before
turning southward and eventually home to Assyria. In these interpretations,
Sargons journey south and back to Assyria along the western shore of Lake
Urmia remains basically the same. The various routes have been generally
labeled the long and short route. Whatever routes he undertook, Sargon
was basically unimpeded, except for the alleged journey to Lake Van (Mus-
carella 2010: 265).
A. Short Route
1. Encirclement of Lake Urmia from the East: The scholars arguing this
route include: Piotrovskii 1959: 105, map; Burney 1972: 138140, 155;
Van Loon 1975: 206207; ilingiroglu 19761977; V. Selim and O. Belli
(Anadolu Arasstirmalari 19761977): 8990; W. Mayer 1980: 2223;
Zimansky 1990; Liebig 1989: 33; Reade 1994: 186; 1995: 38: but to him,
the campaign extended as far as the Caspian Sea; Kroll 1997: 206.
2. No encirclement of Lake Urmia: This view suggests that the campaign
was limited to its western shore area then returned south: Rigg 1942:
132, 135, 138. Rigg was the first scholar to suggest this route. I originally
advanced this view based on observations at Lake Urmia, first in 1960
when travelling down its eastern shore, and reinforced by the discovery
of the Qalatgah inscription (Muscarella 1971:49; 1986. See also Salvini
1984: 4651; 1995: 2829, 92). Zimansky (1990: 5, note 22) states that I
even claim[ed] I was the first to make this observation. Indeed, I did
suggest this, not then knowing the ideas of Rigg 1942, along with those
of others, including Levine 1977. Here (page 150) Levine gives a buried
reference to my 1971 revision of Sargons route, which he dismisses by
describing it as tentative.
B. Long Route
1. Reaching Lake Van: Thureau-Dangin 1912; Wright 1943; Van Loon 1975:
205207; 1987: 260; ilingiroglu 1976/77: 264265; Kleiss 19691970: 131
132, Abb. 3.
528 chapter fifteen
2. Encircling Lake Van: Thureau-Dangin 1912; Piotrovskii 1959: 105; Burney

1972: 155; ilingiroglu 19761977.
My interpretations of Sargons campaign route and the identification of
the sites and districts were first presented in Muscarella 1986, and I have
not amended them. I accept the Short Route, 2: Sargon did not encircle
Lake Urmia from the east, did not reach the Caspian Sea, did not cross the
Zagros from northwestern Iran to Lake Van; he campaigned solely along
Lake Urmias southern and western shores and adjacent districts before
turning back to Assyria. I thus agree with Rigg, Levine, and Salvini. Sargon
made no mention that he encircled the 120 km. long lake, correctly called
by him a sea. As already noted, there are no Urartian sites east of Lake
Urmia, which is recognized by the scholars who proposed encirclement,
viz. Zimansky 1990: 910, 21. Zimansky (1990:14) is aware that non-Urartian
polities existed on the eastern areas of the lake, but claims that there is
no theoretical objection to a march around the lake. However, since the
eastern area was not Urartian, Sargons manifestly targeted enemy, there
is no evidence that he didor would havecampaigned east around the
lake; and, again, if he did so, he would have mentioned it. Van Loon (1987:
259) refers to Sargons mention of a sea twice (lines 233 and 286), and
suggests that two separate lakes are thus involved: Van and Urmia. This
is surely not viable, as Sargon could/would have mentioned Lake Urmia
twice when encountering separate sites close to it. Reade (1995: 3839) even
argues that Sargon reached the Elborz Mountains and the Caspian Sea.
He further believes that Shamshi-Adad V (823811 bc) also accomplished
this feat earlier, thus demonstrating the length of a campaign possible for
an Assyrian army. This is highly improbable; it cannot be demonstrated
from Sargons letter. And the mention of The Sea of the Setting Sun by
Shamsi-Adad is a problem in terms of Assyrian geographical orientation.
Reade assumes it refers to the Caspian Sea, not the mere lake Urmia,
which I believe is not a correct description for that Sea (cf. The Dead Sea).
The many Assyrian artifacts recovered at Hasanlu Period IVB confirm that
Assyria was in contact, most probably by diplomatic/trade activities, with
Lake Urmia in the 9th century bc (Muscarella 1980: 200202, 210217). The
8th Campaign was indeed a long oneas noted here, it continues into the
21st century ad!
Bibliography
Burney, C. (& Lang, D.M.), 1972, The Peoples of the Hills, New York. ilingiroglu,
A., 19761977, The Eighth Campaign of Sargon II, Jahrbuch fr keinasiatische
Forschung (Anadolu Arastirmalari) 45: 252269.
Fales, F.M., 2003, Evidence for West-East Contacts in the 8th Century bc: The Bukan
Stele, in Continuity of Empire (?) Assyria, Media, Persia, G.F. Lanfranchi et al eds.,
Padova. Pp. 131147.
Kleiss, W., 1969/70. Zur Ausbreitung Urartus nach Osten, AMI 19/20, pp. 125
136.
, 1977, Alte Wege in West-Iran AMI 10, pp. 137151.
Kroll, S., 1997, Review of Salvini 1995, in Die Welt des Orients XVIII, pp. 203208.
Lanfranchi, G.B., 2003, The Assyrian Expansion in the Zagros and the Local Ruling
Elites, in Continuity of Empire (?) Assyria, Media, Persia, G.F. Lanfranchi et al eds.,
Padova. Pp. 79118.
Levine, L., 1977, Sargons Eighth Campaign, in Mountains and Lowlands: Essays in
the Archaeology of Greater Mesopotamia, Bibliotheca Mesopotamia Vol. 7, pp. 135
151.
Liebig, M., 1989, Zur Lage einiger im Bericht ber den 8. Feldzug Sagons II. von
Assyrien gennanter Gebiete, ZfA 81, 1, pp. 3136.
Mayer, W., 1980, Sargons Feldzug gegen Urartu714 v. Chr., MDOG 112, pp. 13
32.
Meissner, B., 1922, Die Erorberung der Stadt Ulhu and Sargons 8. Feldzug, ZA,
pp. 113122.
Muscarella, O.W., 1971, Qalatgah: An Urartian Site in Northwestern Iran, Expedition
13, 3/4, pp. 4449.
, 1980, The Catalogue of Ivories from Hasanlu, Iran, The University Museum,
Philadelphia.
, 1986, The Location of Ulhu and Uise in Sargon IIs Eighth Campaign, JFA 13,
pp. 465475.
, 2010, Hasanlu and Urartu, in Bianili-Urartu, J. Reade and S. Kroll eds.,
Munich. Pp. 265279.
Pecorella, P. & Salvini, M., 1984, Tra lo Zagros e LUrmia, Rome.
Piotrovskii, B.B., 1959, Vanstoe Tsarstvo, Moscow.
Reade, J., 1995, Iran in the Neo-Assyrian Period, in Neo-Assyrian Geography, M. Liv-
erani ed., Rome. Pp. 3142.
, 1994, Campaigning Around Musasir, in Anatolian Iron Ages 3, A. ilingiroglu
and D. French eds., Ankara. Pp. 185188.
Rigg, H.A., 1942, Sargons Military Campaign, JAOS 62, pp. 130138.
Salvini, M., 1984, in P. Pecorella and M. Salvini, Tra lo Zagros e LUrmia, Rome.
Salvini, M., 1995, Geschichte und Kultur der Urarter, Darmstadt.
Thureau-Dangin, F., 1912, Une Relation de la Huitime Campagne de Sargon, Paris.
van Loon, M., 1975, The inscription of Ishpuini and Menua at Qalatgah, Iran, JNES
34, 3, pp. 201207.
, 1987, Review of Pecorella and Salvini 1984, BibOr XLIV, 1/2: 252262.
Wright, E.M., 1943, The Eighth Campaign of Sargon II of Assyria (714BC), JNES 2,
pp. 175186.
530 chapter fifteen
Zaccagnini, C., 1981, An Urartian Royal Inscription in the Report of Sargons Eighth
Campaign, in Assyrian Royal Inscriptions: New Horizons, Orientis Antiqui XVII,
F.M. Fales, ed. Pp. 259295.
Zimansky, P., 1985, Ecology and Empire: The Structure of the Uraratian State, Chicago.
, 1990, Urartian Geography and Sargons Eighth Campaign, JNES 49, no. 1,
pp. 121.
Section Two
Anatolia
chapter sixteen
NEAR EAST INVITED REVIEW:

KING MIDAS TUMULUS AT GORDION*
Three Great Early Tumuli. The Gordion Excavations, Final Reports, Volume I.
(1981) R.S. Young, with contributions by K. DeVries, E.L. Kohler, J.F.
McClellan, M.J. Mellink, and G.K. Sams. E.L. Kohler, Editor. University
Museum, Philadelphia. 310 pp., with Appendices, illustrated with figures and
plates. $75.
Long awaited, and after several missed deadlines (the last being Decem-
ber, 1981) the publication of the architecture and contents of the three
largest and richest tumuli excavated by Rodney S. Young at Gordion has
now appeared (July, 1982). It will be immediately noted here that the vol-
ume is one of the most important and significant publications of excavated
material from a major Near Eastern site to have been published in years. The
vast amount of material systematically presented in numerous photographs,
plans and drawings, and in interpretative textual commentaries, revealed in
a detailed manner never before attempted, will surely generate among stu-
dents a renewed interest in the culture and history of the Phrygians. Young
died (October 25, 1974) before he was able to organize for publication his
manuscript with photographs, write the essays and conclusions he planned,
and express all the ideas he kept in his fertile mind. A committee was there-
fore organized in 1975 by F. Rainey, then Director of the University Museum
to collate the material in hand, to supply technical data and to contribute
and solicit appropriate essays. The committee was also charged with orga-
nizing the topics and schedules for future publications of the 17 Gordion
campaigns (19501973); this work is now in progress and soon (one cau-
tiously hopes) we shall see a volume on the lesser tumuli (E.L. Kohler) as
well as a detailed study of Phrygian pottery (G.K. Sams).
In this first volume the committee has accomplished the difficult task
of gathering and organizing the scattered and unpaginated notes of Young,
* This article originally appeared as King Midas Tumulus at Gordion, The Quarterly
Review of Archaeology (1982): 710.

534 chapter sixteen
printing them in the sequence he desired, and (so we are informed) leav-
ing them intact and unchanged with proper concern to preserve his ideas
and style. Aside from assembling the necessary photographs, plans, and
drawings and adding appendices on technical subjects, some of the com-
mittee members, along with two invited scholars (gks, jfmc), have writ-
ten entries and essays on various subjects, including the final Conclusions
chapter. Thus, the volume is the work of several scholars, indeed primar-
ily of Young, but with important contributions by the others, all intimately
familiar both with the excavations at Gordion and with the mind of its
excavator.
Following the wishes of Young, the three tumuli (p, mm, w) are published
in the order of their excavation, not in the chronological order perceived
by the committee. Narration of the excavation problems and strategy, tomb
description, and a catalogue of most of the contents of Tumulus p and
mm are by Young, with further discussion of the pottery by gks, the wood
remains and bronze quadriga by elk, the iron by jfmc, and the Egyptian
Blue and paste by mjm. At the time of Youngs death he had completed
a catalogue of the finds from Tumulus w, but not a description of the
excavation of the tomb and its contents in situ. This latter work has now been
accomplished by kdv, based on Youngs field notes and published writings;
there is also a discussion of the pottery of w by gks. A separate chapter
(IV Commentary) includes discussion of selected groups of objects found
in three tumuli, bull cauldrons (kdv), omphaloi (mjm), belts (elk), pottery
(cks), wood (elk), and bronzes by Young. A final chapter, Conclusions,
is written by mjm. Analyses of wood, food, bronzes, and textiles from the
three tumuli, as well as the latest C-14 dates and a discussion of Phrygian
inscriptions are given in the appendices.
Given the situation that there is a total of six authors in the main body
of the text, it would have been helpful to the reader if each non-Young
entry had been identified by initials alongside the section heading, rather
than with an asterisk matched at the bottom of the page with the authors
name. Because these entries are interspersed within Youngs in the Tumulus
p and mm chapters, too often I discovered that I was unaware that a shift
had occurred from one author to another (for example pages 219 to 233).
In the appendices the names of the contributors are properly supplied at
the heading, as are also the dates when their manuscripts were submitted
(from 1959 to 1979). But dates are not supplied for the time of completion or
submission either of Youngs or the committees contributions.
This multiplicity of contributors has also resulted in the inevitable sit-
uation that there is at times a lack of uniformity in discussions of the
near east invited review: king midas tumulus at gordion 535
same material, in emphases, and even conclusions. Some of the problems

arising from the presence of different authors, writing at different times
and with different perceptions and perspectives, could have been avoided
by stricter editorial control, with concern for consistency in terminology,
and by the inclusion of cross-references or notation that certain unavoid-
able contradictions or disagreements exist. Thus, speaking here of termi-
nology, the city mound is referred to by a variety of terms, never defined,
so that even a student familiar with the Gordion excavations will be con-
fused as to just what is meant, and what level or phase is called to our
attention. Between pages 47 and 49 we are subjected by one author to the
following references to stratigraphy in the city mound: destruction level,
burnt level, early levels, early city, early city settlement, early Gor-
dion, earlier settlement contexts, later contexts of the city, and pre-clay
context; the same author on pages 176, 253, and 256 refers to the early
city, city proper, and early city proper, and on page 216, n. 27 to Pre-
destruction evidence. Other authors refer to the city mound as the hyk
(56), citadel (221, 268, 270, 272), or the city mound (202, 224, 227, etc.).
Except for destruction or burnt level the other terms are vague and non-
informative in the contexts presented, and result in a lack of specificity, not
a minor issue indeed. A general introductory essay, even a brief summary,
on the city mound excavations, with appropriate reference to levels and
phases of Phrygian Gordion, would have been helpful to the reader. Such
an essay would also have established the cultural and chronological milieu
within which the tumuli could be meaningfully situated; its lack is unfortu-
nate.
Turning now to the other issues raised, on page xxix, the Preface (un-
signed) states that the committee members are responsible for the Con-
clusions chapter, but on the very same page we are equally informed that
Machteld Mellink wrote the Conclusions, and indeed, that chapter bears
her name. On page 101 Young refers to the inlaid wood panels as throne
supports while on pages 64 and 261 he refers to them as throne backs or
simply screens. The contributors cite them as screens. Does page 101 rep-
resent an unedited, earlier writing of Young? If so, we should have been
informed, but the throne reference is left hanging, and I suspect that
some readers will wonder about this remarkable object that disappears.
On page 248, n. 128, Young says that only fibulae, not the other bronzes,
have a zinc content, but both A. Steinberg and W.J. Young (289 f., n. 33)
document that at least two bowls tested are also brass. And while Young
claims that the zinc was an accidental addition, Steinberg (288) believes
it is not. Did Young write his comments before Steinberg? Will testing
536 chapter sixteen
continue to clarify this very important issue and to determine if other Gor-
dion bronzes are actually brass?
A few more minor contradictions could also have been noted. On page 156
Young refers to 37 fibulae on the Tumulus mm bier, of which ten had disin-
tegrated, while on page 169 the numbers 32 and 5 are given respectively. On
page 114 Young mentions ten studded belts in mm, where there are in fact
nine (147ff.), the tenth being of a different type and all bronze. And while
kdv says (221) correctly to my mindthat there is no Urartian material at
Gordion, mjm (268) speaks, with no evidence given, of Urartian influence
there (see also Mellink 1979, 256).
Although it is maintained (xxix) that an updating of the notes of Young
was supplied, the documentation presented does not support this claim,
and one is disappointed by the lack of both relevant bibliographical ref-
erences and discussions of comparanda, not only in Youngs entries but in
some of the contributors as well. For example, there is not a single com-
ment concerning the techniques of tumuli construction elsewhere in Ana-
tolia, nor in the Near East and Europe. Indeed, a separate essay devoted to
this subject in a volume on Phrygian tumuli is not too much to expect, all
the more so as Young basically neglected this aspect of Phrygian archae-
ology. That a number of excavations have recorded the various methods
of tumulus filling and the placement of the underlying tomb off center is
ignored (e.g. on Cyprus, at Sardis, Kerkenes Dag, Ankara, and in Iran at S
Girdan and Takht-i-Suleiman: for a summary see MMA Jour 2, 1969, 22 f.,
MMA Jour 4, 1971, 25f.). In the wood catalogue for Tumulus p we miss a
discussion of the ivory carvings from the city mound at Gordion, as well
as reference to those at Nimrud, Hasanlu, and elsewhere, with pertinent
comments on the relationship of ivory and wood carvings. Nor is there a
mention of the preservation of wood sculpture at other Near Eastern sites
at Ebla, Nimrud, Hasanlu, Samos, Karmir Blur, the Altainor to the pos-
sible reasons for their preservation in each instance. Formal parallels are
given for one wood sculpture (54f.) related to Iranian work, but the cited
examples are all Achaemenian gold, and about 200 years later in date (also
note that although attributed to Hamadan, they are not excavated pieces).
That Egyptian Blue artifacts (268) and animals headed vessels (121 ff.) were
excavated at Nimrud and Hasanlu, that vessel attachments in the form of
birds occur at Hasanlu, is nowhere mentioned. Further, that parasols are
represented, aside from Ehnali, on Assyrian and Achaemenian reliefs in
association with royalty, that fans are commonplace in Near Eastern rep-
resentations of banquet scenes (74ff., n. 145, 146), and that there is recent
work published on Phrygian fibulae in the West (K. Kilian, 1975, 151 ff., and
E. Sapouna-Sakellarakis, 1978, 120ff.: Prhistorische Bronzefunde xiv, Band

2 and 4, respectively) is also not revealed in the volume. Further, there
are no references to the objects from the tumuli previously published else-
where.
Is one asking too much? Will it be argued that this information will be
furnished elsewhere? But where else other than in a final report is a full
discussion to be expected, one that not only concerns itself with what was
excavated at Gordion, but how this material relates to what is known at
nearby and distant sites? In this context concerning the commentary on
bull cauldrons by kdy (219ff.) let it be recognized once and for all that the
Copenhagen bull cauldron may not, and cannot, be accepted as an exca-
vated object from Cumaeor any other place. The discussion of pottery by
gks (251ff.) and the fibulae by Young (239ff.), are fairly good summaries with
more or less relevant bibliography. The committee was concerned solely
with the finds from the tumuli, not with the presentation of an integrated
view of the culture of Gordion and its historical/archaeological role in the
Near East.
Yet, the above comments notwithstanding, there are no reservations or
hesitations in my claim that we have in this volume a fundamental work
of Anatolian archaeology. Dramatically, it presents under one cover a large
quantity641 objects are cataloguedof excavated artifacts composed of
various materials and made in a variety of techniques, that collectively
illustrate as never before, the creativity of the Phrygians and the position
of Gordion as a major political and economic center in the first millennium
bc. This documentation is accomplished by the Editors decision to publish
each object with a catalogue entry and photograph (only a few duplicates
are not photographed; and that some photographs are too small for full
appreciation is unfortunate but economically understandable), as well as
a generous 148 fine drawings and plans, all set in an excellent format and
binding.
Aside from the insights allowed of the enormous wealth and power of
the Phrygians in the late 8th century bc, as evidenced by the large quan-
tity of locally made and imported material and implied by the existence
of the large labor force we recognize was required to build the tumuli, we
are unreservedly awed by the skill of their craftsmen, remarkable even by
Near Eastern standards. Phrygian bronzeworkers were capable of manu-
facturing a large polythetic set of bronze/brass vessels of different shapes
and varieties, fibulae of different types and intricate construction (one must
read Youngs description of the construction of a fibula on page 249 to get
a good insight into Phrygian craftmanship and ingenuity). They combined
538 chapter sixteen
leather and bronze belts, and employed different techniques of hammer-

ing, casting, or a combination of these techniques on one object. Their
woodworkers painstakingly constructed delicate screens with minute inlaid
patterning, complex tables and parasols, and more mundane tables, plates,
and so forth, all with consummate care and skill in turning and bending.
They also sculptured superb animal figurines. Their weavers made multi-
colored patterns from a variety of materials, and the preservation of textiles
archaeologically supports the ancient reports about the quality of Phrygian
weaving. Further, their engineers and architects planned and built sturdy
wooden tombs with neat joining, anticipating and resolving the problems of
stress that would result from many tons of stone and clay that would overlay
them.
These accomplishments are revealed to the modern scholar through
the careful and detailed descriptions of the material by Young and his
colleagues. They have also identified specific artifacts that may legitimately
be called Phrygian, as opposed to other Near Eastern attributions, such as
besides the already well-known fibulaebronze ladles, petalled omphaloi,
bowls with bolster and rim bands, belts, and T-shaped attachment plates,
not to mention the woodwork.
The contributors to the appendices furnish valuable information that
fleshes out the catalogue descriptions. For example, A. Steinbergs cautious
essay on metal analysis presents the data objectively; it raises questions but
gives non-categorical answers. R. Ellis equally cautious comments on the
textile remains are clearly written, and he too refuses to make facile identi-
fications of the material studied. The essay by Cl. Brixhe on the inscriptions
from Tumulus mm is also valuable, but more with regard to his summary of
previous work than to the present position of the problem. His dating (275)
of the importation of the alphabet to Phrygia in the late 9th or early 8th cen-
tury is based on subjective internal analysis, rather than on archaeological
grounds, and is unacceptable. Aside from an inscription found in the city
mound under a floor (Hesperia xxxviii, 2, 1969, 257, no. 29), which Young
suggested is to be dated about 750 bc or perhaps earlier (or later?), none
of the tumuli except mm have inscriptions (and its date is still not resolved,
infra): 750 bc, if that early, remains the upper limit for the introduction of the
alphabet at Gordion, or elsewhere in Phrygia (and it may well be a decade
or more later; see also H. ten Cate 1967, 121).
One of the most intriguing sections of the volume is Youngs narration
(a better word here than description) of the excavation process of tumu-
lus mm, completed in 1957 but preceded by two campaigns of drilling, in
search of the elusive tomb within the enormous mound (53 meters high and
ca. 300 meters in diameter). After the drilling period it took a further two
full months of digging, half of this time in 24-hour-a-day shifts, involving
first opening a trench, and then digging a tunnel for a combined length of
135.20 meters, to reach the tomb. It was discovered to be surrounded imme-
diately by a revetment log wall and a square stone wall three meters in height
outside that, all covered with many tons of stone rubble. A gabled structure
made of neatly cut pine wood timbers, the tomb was intact and uncollapsed,
in itself an historic discovery. Inasmuch as there were no windows or doors,
an opening was cut through the west wall, and upon entering one found one-
self directly before the bier holding its occupant. For the record, Machteld
Mellink was given the honor of being the first non-Phrygian to enter the
tomb (although I am informed another staff member improperly made that
claim).
Youngs text is, typically, clear and concise, yet at times difficult to follow
primarily because of the problems inherent in describing such a complex
tomb structure. And often there is inadequate cross-referencing and coor-
dination between the text and an appropriate figure or plan (presumably all
references to figures were added after 1974). Thus, for example, figures 61,
63, and 64 better illustrate mortising techniques than do the cited figures 54
and 59 (88). The important figures 5262 (executed by C.K. Williams ii)
require internal labelling in order to identify details mentioned in the text
e.g. the arrows in figure 62and they are not sufficiently cited in the text.
(Figure 63 is a masterful isometric drawing of the whole tomb complex
by C.K. Williams ii, but we selfishly miss an isometric drawing of internal
structure of the tomb itself.) Certain features of the initial building of the
tumulus described by Young (80), that there was a successive line of stones
laid to determine the center and that there was a laying down of a ring of
filling [clay] around a predetermined perimeter for the tumulus, are not
indicated in the plans and sections (figures 5052). Moreover, the section
drawing of figure 52 is merely a sketch, not an empirical drawing; it was
copied in Philadelphia from Youngs notebook. No sections are supplied for
Tumulus p and w. Further, a section drawing of Tumulus p (figure 3), shows
a pit although it is not mentioned in the text, and the description of the
excavation suggests that such a feature could not have been noted even if
present (note that in AJA 61, 1957, 325f., there is mention of a pit). Tumulus
w had a tomb pit, but here too the section drawing does not show it (fig-
ure 114).
On page 263 Mellink casually notes that Youngs description of the enter-
prise [the excavation of Tumulus mm] allows the reader to read between the
lines that the discovery was a rare confrontation with an intact part of the
540 chapter sixteen
past Between the lines presents the problem precisely. For although
the description takes the reader step by step in the long and arduousand
courageousjourney that led to the discovery of the tomb, one is aware that
it consciously lacks any hint that the excavator and his team may have expe-
rienced personal excitement. While the description is correct, it is written
with an intent to avoid any expression or pride and joy in a major discov-
ery. A rare confrontation with the past. Alas, it is an attitude that also
obtained during the excavation itself.
The most important conclusions expressed in the volume concern the
relative and absolute chronology of the tumuli, in particular the absolute
dates assigned to Tumulus w and mm. These conclusions are issues of con-
siderable importance and will keep scholars pens occupied for some time.
Indeed, they crucially affect not only the internal history of Gordion itself,
but of Phrygia at large, and equally the dating of the Phrygian alphabet, the
dates for the transmission of Phrygian material abroad, and, not least, the
nature of archaeological interpretation. With these issues in mind, the fol-
lowing comments are offered to initiate discussion.
Young expressed the belief in all his previous writing, and never changed
his mind, that the three tumuli, along with Tumulus iii, excavated in 1900
by the Krte brothers, were pre-destruction (ca. 696 bc) in date. To him,
w was the earliest, and he dated it to the end of the 9th century bc or
slightly later (199). mm was next in date, about six or seven decades later,
ca. 725/720717 bc, or the latter half of the eighth century (102, 109, 232),
followed by iii and p in time: thus, the sequence w, mm, iii, p, with a range
from ca. 800 to 700bc or a full century. The contributors hold a different
view, with regard to both the relative and the absolute dates. While they
all share Youngs view that w is the earliest, they believe that mm is the
latest in the sequence, and it seems that some believe further that mm was
not built until after the Kimmerian destruction, which we must presume
dates it ca. 696/5 bc. In this sequence we have w, iii/p, mm (176, 198 f., 215 f.,
233ff., 236ff., 239, 254, 263f., 266, 269ff.); the relative sequence of iii and p
is unclear, and they are considered to be too close in time to separate them
chronologically.
Turning first to Tumulus w, we note that kdv (199) dates it decades if not
generations before the catastrophe, which I interpret to mean (assuming
that if not means even) that w is either ca. 760 or 790 bc (accepting
the conventional 30 years for a generation). Equally ambiguous is Mellinks
suggested chronology, which implicitly reveals uncertainty. On the one hand
(264, 270) she seems to accept a ca. 800bc date, in part based on the dating
of a bronze vessel, w9, that Young originally associated with one excavated
in tomb 30 at Assur bearing the inscription Assurtaklak (of which more

below). On the other hand (272), she later dates w to the generation before
Midas, the beginning of whose reign is calculated by her (271) as occuring
between 735 (infra) and 717 bc; Eusebius date of 738bc is not mentioned in
this context. In this chronology, w is thus dated at the earliest to ca. 765 bc,
i.e. a generation after 800bc, or to ca. 750bc (and note that in 1979, 252,
Mellink claims that w is slightly earlier than mm). The other contributors
avoid assigning an absolute date to w.
The conclusions of kdv and mjm that w is the earliest of the Gordion
tumuli are based on stylistic analyses of the artifacts that seem more than
not to be justified, but which still require more study. They see an evolution
(198ff., 233ff., 264, 266ff.) in the manufacture of fibulae (especially types
xii, 3, 7a and 7, and 13), omphaloi (size of bowl and omphalos and the
use of ridges petals) and the presence or absence of certain pottery forms,
and the technique of tomb construction, collectively progressing from w
to iii/p to mm. The trajectory they see is basically, and classically, from the
simple to the elaborate. However, in the discussion of the belts from w, p,
and mm, this perceived art historical development is conveniently turned
on its head, and we are informed matter of factly that the more elaborate
belt from w seems ancestral (236; italics added) and is obviously earlier
(239) than the simpler (237) belts from m, which are a survival. This, I
am afraid, is a forced conclusion based on an a priori assumption that w is
earlier than mm, and it contradicts the sequence stages suggested for the
other material. Equally forced I suggest, is the stylistic discussion of the
wood plates and trays (238f.), where the sequence w, iii, mm is explicitly
accepted as a given, leading to observations on what are innovative forms
and innovations.
While I do not consider it improbable that w is the earliest tumulus
under review (a conclusion I tentatively presented in 1967, 3, 43, 46 f., with
very little information available to me at the time), I do seriously doubt
that it is to be dated close to 800bc, or even within a decade or two of
that time. Whatever its absolute date, it cannot have been built by much
more than a few decades, if that much, before the construction of mm
at the earliest ca. 750/740 bc. For, while the contributors have emphasized
stylistic differences among the tumuli finds, they have not spoken to the
similarities, which is fundamental in a discussion of chronology. The sole
exception is gks who cogently notes that the pottery of w, iii and p, and
also less so mm, could not have been made far apart in time (215, 252 f.,
256f.), and he makes the telling statement (referring to the analyses of kdv)
that The relative earliness of Turmulus w manifests itself more readily
542 chapter sixteen
through its bronzes than through its pottery (215). Parallels in form and
type among the finds from tumuli include (and I give them at length to
document my position; the letters and numbers stand for the tumulus and
catalogued object): bronze spouted vessels (w5, p7278, iii, mm14); pottery
spouted vessels with stepped interior spouts (w61, p75); geometric painted
vessels (w61 [is this not the most developed of all?]; p4953, iii); painted
panel vessels (p5557, iii: on page 49 ks says the panels of an example from
the city mound [pl. 96, b] are small and relates them to iii; on page 270
mjm implies it has larger panels than iii); monochrome spouted vessels
(w63, p7278, iii); leather belts (w25, 26, iii, mm170179); bronze belts (p34
36, mm180); type xii, 7a fibulae (w iii, also city mound destruction level);
bull cauldrons (wi, 2, mmi, 12, 13); oinochoe base with a pin embedded in
wax (w6, mm27); animal headed finials (w8, p36); bronze ladles (w, iii, p,
mm); omphaloi (w10, p11; and compare w11, p12, mm125130; w1214, p14
28, mm131167; compare also the plain bowls w23, p2730, mm168169);
pottery amphorae (w6572, p91104, mm372377); wood screens (w80, p151,
mm378, 379); wood utensils (w7579, p121154, mm380387). w and mm
share ten of the above parallels, p and mm share seven, p and w share
thirteen. The main differences between w and mm are the lack of type xii, 7a
fibulae and painted vessels in mm, and the lack of dinoi in wall negative
evidence.
External evidence does not help in the dating of w. Although the often
cited bowl from Assur bearing the name Assurtaklak, the same name as
the Assyrian eponym of 806bc, has been brought forth in discussions of
ws chronology (198, 234, 270), it may not in fact be so employed. In the
first place, as Mellink has noted, its shape is not the same as w9 to which
it has been compared. In the second place, it is quite possible that the
vessel is almost a century later in date than 806 bc. Tomb 30 at Assur also
contained a fibula, the date of which cannot be earlier than the late 8th
century and could be 7th century, and a lamp of a type excavated in a late
8th early7th century bc burial at Baba Jan in Iran (Iran vii, 1969, 125,
fig. 7:1, 2). These artifacts date the deposition of the objects in Tomb 30 a
century or more after 806bc: which suggests either that the bowl was an
heirloom (but visible and in use for a long time), or that another man named
Assurtaklak inscribed it with his name (see also P.H.G. Smith in BaBesch 56,
1981, 10).
Because Urpallu, the king of Tyana, is represented at Ivriz wearing Phry-
gian clothing and a Phrygian fibula of type xii, 9, the exact date when the
relief was carved is clearly of great importance for the dating of Phrygian
fibulae, type xii, 9 in particular, and by extension back in time, xii, 7a.
Akurgal (Die Kunst Anatoliens, 118) dated the relief after 730bc (see also
Orthmann in PKG 1975, 432, um 730 v. Chr.). Young (245, n. 115) placed the
relief earlier in the years around 738bc, more probably before than after
(a position taken by me in 1967, 20). This chronology is based on a reference
to Urpallu in the Assyrian annals of Tiglath-Pileser iii; but other, albeit
complex, evidence for the date of Urpallu now exists. A tablet excavated
at Nimrud written by an unnamed Assyrian king mentions both Midas
(Mita) and Urpallu as contemporaries (J.N. Postgate, Iraq xxxv, 1973, 21 ff.).
Postgate firmly dated the letter to the time of Sargon, to 709bc specifically
(agreeing with Saggs, Iraq xx, 1958, 205ff., who first published the letter).
Mellink (371; see also 1979, 249ff.) is inclined to date the letter to the time
of Tiglath-Pileser iii, to ca. 735 bc (following H. ten Cate 1967, 122), and she
believes that the relief at Ivriz was carved at or close to the same time.
Inasmuch as the dating of the relief would supply independent evidence
about the time when type xii, 9 fibulae were in use, and if, as plausibly
argued by kdv and Young, this type is a development of type xii, 7, itself
a development of type xii, 7a, then the latter two types would have been
created sometime earlier than the relief. Thus, if the Ivriz relief was carved
in the 730s, we would have a date in that decade for the floruit of all three
types, and equally a projected date back in time of some undetermined years
for the first appearance of types xii, 7 and 7a. But if the Assyrian letter, and
concomitantly the relief, is from the late Sargonid period, from ca. 709bc,
we would then have a contemporary date for the three types in the previous
decade, and probably in the 720s, even the 730s, for types xii, 7 and 7a.
In short, the Urpallu relief is of potential, but at present limited, value for
determining the earliest chronology of the three fibulae types discussed, and
consequently for the objective dating of the xii, 7a fibulae found in quantity
in Tumulus w.
I do not think I am wrong to suggest that the attempts to date Tumulus w
close to 800 bc are ultimately based on an underlying assumption (but one
never mentioned) that the archaeological record indicates that Gordion was
an established city by this time. In AJA 1966, 273 ff., and in AJA 1968, 239 f.,
Young presented evidence for the existence beneath the destroyed level at
Gordion of an earlier level with seven phases. One may speculate about
the incipience, duration, and chronology of this level and its phases, very
little of which was actually exposed, but there is nothing in the published
record that compels acceptance of a pre-8th century bc date for the time of
the construction. Nor does the published evidence from the deep sounding
(AJA 1966, 276ff.), which revealed no architecture or discrete levels as such,
compel us to accept the existence of a 9th century settlement at the site.
544 chapter sixteen
Nevertheless, I am aware that all the data from the sounding have not
yet been published and that disinterested scholarship requires an open
mind on this fundamental problem. Therefore, in the context of the present
discussion one may only conclude (patiently awaiting enlightenment from
future publications) that even if there was a Phrygian (and n.b., not a pre- or
proto-Phrygian) settlement at Gordion ca. 800bc, it cannot be assumed that
Tumulus w dates to that time; and if there was no such settlement before the
8th century bc (whatever incipient date), then Tumulus w must have been
built subsequent to that time.
Evidence from other Phrygian sites in the east seem to be less ambigu-
ous on the matter of chronology and yields no evidence for pre-8th century
occupation. At Bogazky (Bossert, MDOG 89, 1957, 58 ff.; MDOG 94, 1963,
53ff., 68ff.; Opificius MDOG 95, 1965, 81ff.) excavations indicate an 8th cen-
tury date for the earliest post-Hittite occupation there. And I suggest that
Opificius attempt (ibid., 87) to push the Phrygian pottery sequence back to
ca. 800 bc, based on an incorrect 9th century date assigned to a vessel from
Carchemish, and an equally incorrect date for the fibulae there, is based on
speculation, not empirically determined evidence (see also G.K. Sams, AS
xxiv, 1974, 181). The same situation with regard to the post-Hittite settlement
and chronology obtains equally for Alishar (Akurgal 1955, 37, 155; also K. Bit-
tel, Hallusha, NY, 1970, 138). Indeed, the earliest post-Hittite level at Alishar,
4c, contained both Near Eastern fibulae, of late 8th/7th century types, as
well as Phrygian examples, surely indications of an 8th century occupation,
not earlier.
Except for the possible, and still to be demonstrated, exception of Gor-
dion, no Phrygian site can be convincingly dated before the 8th century bc.
In this respect Akurgals (1955, 24, 112ff.; 1965, 468 ff., 472) short chronology
for Phrygian culture (but not his suggested low date of 725 bc for the found-
ing of Gordion) still holds up; Phrygia had a glorieuse mai trs courte vie.
In addition to archaeologicalinternal evidenceit is against this back-
ground that Tumulus w should also be studied and its chronology evaluated.
We do not know whether the Phrygians constructed tumuli burials imme-
diately after they built their earliest cities, but on the evidence from Tumu-
lus w as interpreted here the earliest Phrygian tumulus recorded up to the
present time does not pre-date the mid-8th century bc.
Until the very last page (272) only one of the contributors offers his
views on the absolute date for the construction of mm. gks (176) sees it as
roughly coeval with the citys destruction, but he also cautions against
too low a dating and suggests that mm is still close to the date of p and
the time of the early citys end. I read this to mean that mm was sealed a
relatively short time before the Kimmerian raid, an opinion I believed all
the contributors held. It therefore came as a nonsequitur that one/some of
the contributors (see page xxix for confusion concerning who is responsible
for the Conclusions attributed to mjm) consider mm to have been built and
sealed after the destruction.
Aside from being a nonsequitur within the context of the volume itself,
the announcement that mm is a post-destruction construction is more sig-
nificantly an archaeological nonsequitur, one that bends the evidence avail-
able and takes it beyond its natural boundaries. However late the tomb
may be, cognizance of and fidelity to the full archaeological record compels
comprehension of a context within predestruction Gordion. The contrary
statement offered with no set of detailed evidence, demands of others a pre-
sentation of proof that it does not have, and that cannot be accomplished.
The impulse behind the chronology as presented is informed by a fiat, and
it floats in the air (272): Preliminary analysis supports the interpretation of
the occupant of mm as Midas. Which announcement automatically raises
the question, what analysis, and why Midas? Indeed, if the buried man is
Midas it follows that the tomb is post-destruction, inasmuch as it is accepted
in the text that Midas killed himselfas reported by ancient historiansat
the time of the Kimmerian attack; the tomb therefore will have been built
after the suicide/Kimmerian attack, not, obviously, before. However, if one
takes a disinterested position with regard to the identity of the entombed
individual, Midas ceases to be a viable candidate. For the chronology of
the tomb alone, determined by both the internal evidence and historical
and economic analyses, will inform us whether Midas is buried therein, not
the assumption that because the deceased is Midas we therefore know the
chronology. (Of course, if the historians are in error and Midas died some-
time before the destruction, Midas could be buried in mm: but since we
cannot know this, we cannot discuss it.)
Mellink herself (270; assuming her to be the sole author of the Conclu-
sions) provides some of the pertinent archaeological evidence for dating
mm in the late 8th century bc (evidence discussed by others, including
Young): the situlae and type xii, 7 fibulae, both represented in Sargons city
at Khorsabad, and the xii, 9 fibulae, represented at Ivriz. The city mound
parallels with Tumulus mm might suggest (with gks) that the latter was
built close to the time of the destruction, but they do not necessarily indi-
cate a post-destruction date. More study is required for a full understand-
ing of the chronological relationship between certain objects found on the
city mound and in the turnuli, But one surely cannot cite only those from
the former that parallel finds in mm while ignoring the occurrences of
546 chapter sixteen
geometric painted pottery, amphorae, and type xii, 7a fibulae found both in
w and on the city mound.
Then there is also to be considered, and not parenthetically, the matter of
the great quantity of the deposited material, including wood and cloth, that
could not have survived the fire recorded almost throughout the remains
on the city mound. Many questions are begged when it is claimed that
the material could have derived from other Phrygian cities because Midas
citadels were not all looted at the time when Gordion suffered destruction.
Not unexpectedly, no specific cities are mentioned, for the truth or error of
this claim defies precise analysis. Most, probably all, Phrygian cities, west
and east, are archaeologically attested to have been destroyed, although it
is not possible to know whether one was destroyed at the same time as, or
much later (months, years?) or earlier than, another. No other Phrygian site,
not even the Ankara tumuli, has yielded evidence to indicate that anyone, or
all sharing the responsibility, could have sent hundreds of precious objects
to Gordion after destruction, let alone even before.
It is further suggested that the actual work involved in constructing the
tumulus and tomb, if not provisioning it, was locally accomplished: The
effect of the Kimmerian raids was not such as to destroy the city of Gordion
or its dynasty. Tumuli continued to be erected in the 7th century, and the
rebuilding of the citadel started gradually under the protection of the mud-
brick fortification of Kk Hyk. But the citadel was destroyed and aban-
doned, with no clear evidence of reoccupation, and the alleged rebuilding
is by no means certain or fully understood, nor did it gradually or other-
wise reach fruition. Further, while the Kk Hyk (the excavation and
history of which is not discussed) was built sometime after the destruction
of the city mound, we do not know how much time elapsed before work
commenced, or whether it was solely a fort or a royal habitation. And here
too questions are begged, questions that are neither petulant nor gratuitous.
Did the alleged survivors at Gordion include architects and engineers who
were capable of planning and supervising mms monumental construction,
the likes of which did not exist earlier anywhere in Anatolia, and only later
in Lydia? Where did the work force come from, the hundreds of men and
women required to build the tomb and to carry the countless tons of rubble
stone and clay filling, individuals who were free of other tasks, including pro-
tecting their (rebuilt?) homes from further Kimmerian raids? What secure
authority had the compulsion and the finances to both assume control of
the enterprise, with its problems of organization, payments, housing and
feeding, the gathering of timber, and to command an army to protect the
work force as well as the objects transported from other cities? And where
were the Kimmerians and why would they allow such booty to go unplun-
dered? Here, with Sherlock Holmes dog in mind, it is not an inconsequential
matter to note that it would have been a very curious incident indeed if
they did nothing, either being present or having recently passed through
Gordion, In short, the political, economic, and military forces and conceptu-
alizations implicit in the projects planning and construction phases cannot
be perceived as entities existing in the chaotic post-destruction period of
Gordions history.
Mellinks opinion concerning the late date of mm and the identity of its
occupant as Midas is not a new one, for it has been argued by E. Akurgal
since 1959 (although his views on this matter are not once mentioned in
the text). This position has the support of other scholars such as Snodgrass,
Karageorghis, and Hrouda, but not of Boehmer, H. ten Cate, H.-V. Herrmann,
Lloyd, Calmeyer, and Burney, as well as the reviewer, none of whom are cited
in the text. Rodney Youngs views on the chronological placement of mm are
summarized on page 102; he did not believe that a tumulus on this scale
could have been made immediately after the Kimmerian catastrophe. He
thought the tumulus belonged to Midas predecessor presumably named
Gordius. This reviewer continues to agree with Young against the contrary
views of the contributors, A major Phrygian individual, probably a king, was
buried in mm by a rich and powerful state in its prime, during a period of
security and peace, not chaos. The time of the deposition will have been
some years before the destruction, perhaps even before the Kimmerians
were a potential threat, although on this particular issue we cannot speak
with certainty. In any event, the exact date could be close to the time
suggested by Young, or a decade or more later, sometime between the
years 720 and 700bc. Unfortunately the individual remains anonymous,
although his name could be Gordius. And it is probable that the impelling
force behind the enterprise was King Midas who, by erecting the largest
tomb and tumulus built up to that time, attempted and succeeded not
only to honor a revered relative but also to impress and overwhelm his
neighbors, friends and enemies alike, Sargon, Urpallu, and the North Syrian
and Urartian kings. The tomb and its overlying tumulus, I suggest, was
in addition to its sepulcharal function a self-conscious Midian-Phrygian
statement, an artifact cum message, a symbol of Midas political authority
and canniness. To this extent, and only to this extent, may Midas be brought
into a discussion of Tumulus mm, which letters signify not Midas Mound
but Midas Mound: a tumulus that will be identified in the minds of many
of us as an eternal monument not only, to Midas, its postulated builder, but
equally to Rodney S. Young its historical excavator.
548 chapter sixteen
(I was able to recognize a few typographical errors: page 4, left: figure 2

= 3; page 24, right; figure iiia = 14a, b; page 148, left: mm193 = 173; page 179,
left: (pl. 44c) = (pl. 44b); pl. 88c = w4; pl. 8ac = p3: figure 101 was drawn by
Grace Freed Muscarella and figure 117c by A. Seuffert.)
On the day of resurrection all animals and

inanimate things will be given the power
of speech (Gwendolyn MacEwen)
References Cited
Akurgal, E. (1955) Phrygische Kunst, Ankara.

. (1959) Chronologie der phrygischen Kunst. Anatolia IV, pp. 115121.
. (1965) Les Problems de lArt Phrygien, in Le Rayonnement des Civilisations.
Paris, pp. 467474.
Houwink ten Care, H.J. (1967) Kleinasien zwischen Hethitern und Persern, in
Fischer Weltgeschichte, Band 4, 112134.
Mellink, M. (1979) Midas in Tyana, in Florilegium Anatolicum, offert E. Larouche,
edited by E. Masson and Ch. Brixhe, Paris, pp, 249257.
Muscarella, O. White (1967) Phrygian Fibulae from Gordion, London.
chapter seventeen
THE IRON AGE BACKGROUND TO

THE FORMATION OF THE PHRYGIAN STATE*
Abstract: Recent excavations at Gordion have revealed below the destroyed Phry-
gian city (ca. 700bc) an early Iron Age settlement with handmade coarse ware,
which is followed by a settlement that contains the earliest Phrygian pottery forms.
The handmade ware relates to that from Troy and the Balkans and is considered
firm evidence of the historically recorded migration of the Brygians into Anatolia.
A suggested chronology for the two early settlements is posited, based primarily
on information from Troy. This chronology is then examined together with the
information derived from preserved ancient traditions. A hypothesis is generated
regarding the chronology of the establishment of kingship in Phrygia. This event is
posited to have occurred in the late ninth century bc, and the historical King Midas
is considered to have been the fourth Phrygian king to reign.
From the modern scholarly perspective, the Anatolian Iron Age is said to
begin with the destruction and collapse of the Hittite and neighboring
states, in the years around 12001180bc, a chronology derived from texts
from Bogazky, Ugarit, Meskene-Emar, and Egypt (for a summary of the evi-
dence see Hoffner 1992; Yon 1992; Caubet 1992; also Astour 1965; Freu 1988;
Otten 19761977; 1983). Destructions of the Mycenaean centers occurred
about the same timeand may have resulted from some of the same causes.
In both regions thereafter writing ceased to exist until the eighth century bc,
giving rise to the alternative term Dark Age for the period (e.g., Akurgal
1983; Muscarella 1988: 417). But whereas in the Aegean during the postde-
struction centuries, artifacts, burials, and architecture are archaeologically
documented, indicating the continuous presence of a population, material
remains are lacking in most areas of Anatolia. Such a condition signifies for
Anatolia a darker Dark Age and a less comprehensible cultural and histori-
cal situation than that which existed in the west.
Burned or abandoned settlements have been recorded at practically
every Late Bronze Age site excavated in central, southern, and eastern Ana-
tolia (Akurgal 1983: 7578; Bittel 19761977: 3943, 4849; 1983: 2628, 3035,
* This article originally appeared as The Iron Age Background to the Formation of the
Phrygian State, Bulletin of the American Schools of Oriental Research 299300 (1995): 91101.
550 chapter seventeen
3739; Muscarella 1988: 417419; Drews 1993b: 811, fig. 1). The magnitude
of the political and social disruption is indicated by the fact that not a
single site is recognized to have been directly reoccupied; this process of
resettlement did not begin in most cases until centuries later.1
The identity of the forces that destroyed the Hittite polity still remains
unknownalthough a number of candidates, e.g., the Sea Peoples (specif-
ically named as the destroyers by Rameses III), Muski, Kashka, or alliances
of these powers have been tendered and supported by various scholars. To
date neither archaeology nor textual remains in fact furnish definite answers
or even clues on this issue (Akurgal 1983: 72; Bittel 1983: 27; Hoffner 1992: 49;
Drews 1993b: 4861). While indeed texts from destruction levels at Bogazky
and Ugarit refer to military threats by land and sea against the Hittite and
allied states, it is uncertain if one may attribute them specifically to the
final destructions; and names of enemies for this time have not been recov-
ered (Astour 1965; Otten 19761977; 1983; Hoffner 1992). A text written in
the eighth regnal year of the Egyptian pharaoh Rameses III, ca. 1180/1176bc,
is both the last written reference to the Hittite state and a post quem non
chronological marker for the destructions (sometimes called the Catastro-
phe). It records that a massive force of diverse nationalities coming from
the ocean and islands in the sea invaded both Egypt and the Asiatic main-
land. The invaderscalled by modern scholars Sea Peoplesare said to
have destroyed Cyprus, Carchemish, Arzawa, and Hatti. Rameses text is the
only specific reference to the Late Bronze Age catastrophe known, and it
must reflect some historical reality. This is maintained here even if, as has
been suggested, several years activities may have been conflated by Rame-
ses, or that the Sea Peoples never penetrated into the Hittite heartland but
only weakened the center for others to conquer (Bittel 1983: 4647; Hoffner
1992: 46, 49).
The first historical documents mentioning a polity in Anatolia after the
destructions are from Assyria. In his first year, Tiglath-Pileser I (ca. 1115/1112
1074bc) defeated a people called Mushku with their five kings (chiefs?),
on the western shore of the Tigris River in southeastern Anatolia, a territory
the Mushku had controlled for 50 years. This text documents the presence
1 Some scholars have argued against the existence of a gap in continuity between the Late
Bronze Age and the Iron Age (e.g., DeVries 1987: 6; Gunter 1991: 103, 106; Mellink 1991: 626).
There is, of course, the problem of continuous references in Assyrian texts to the Mushki
living somewhere in central or eastern Anatolia (Muscarella 1988: 417; Laminger-Pascher
1989: 16, 1920). See below for discussion of the gap at Gordion.
the iron age and the formation of the phrygian state 551
of Mushku in eastern Anatolia since at least the 1160s bc, that is, within two
decades (or less) of the Late Bronze Age destructions. Since they are never
mentioned in Hittite texts, it is plausible that the Mushku arrived in Anatolia
during or not long after the time of the destructions (and it is plausible to
ponder, but not to assert, that they either may have played a role in the
destruction of Hatti and neighboring polities or entered the area after the
destructions when there existed no power to resist them).
More than 200 years later the Mushku appear briefly in Assyrian texts,
still mentioned as enemies. In the reign of Sargon II (722704bc), over 150
years later, they are mentioned again, now called Muski, and ruled by a king
named Mita.
Mita of Mushki occurs as such only in the Sargon texts,2 but it is most
probable that he is the king Midas of the Phrygians mentioned in Phrygian
inscriptions and by Greek historians, first by Herodotus (VII.73; Muscarella
1988: 417; 1989: 333334; cf. Laminger-Pascher 1989: 1617, 2324, 38, n. 55
who vigorously rejects the equation). The ancient historians record that
Midas reigned in Gordion, on the Sangarios River, that is in westernnot
easterncentral Anatolia. If, as seems manifest, Mita is the Semitic variant
of the Phrygian name, then the appellations Phrygia and Mushki denote
the same polity. This dual nomenclature might signify that the king ruled
over two originally separate groups/peoples (as, e.g., Mellink 1965: 318320),
the western one better known to the Greeks, the eastern to the Assyrians.
Whether the alleged two groups, if such were the case, were originally
related is presently unknown. Furthermore, the appropriate Phrygian form
Midas, not Mita, occurs in Phrygian inscriptions not only in the west but
also east and south of the Halys River (but not at Babylon [as Mellink 1965:
320]), which may or may not have been the (original) Phrygian-Mushki
boundary (Mellink 1965: 320322; 1991: 623, 625). It is difficult to accept the
view that the Assyrians meant anything other than the polity that the Greeks
(and we) call Phrygians when they wrote of their encounters with Mita of
Mushki.
2 Postgate (1973: 23) mistranslates the Assyrian appellation Mita the Mushkian as Mita
the Phrygian, but on p. 24, n. 4, he paradoxically denies the equation of Mushki and Phrygia;
Hawkins (1982: 421) also mistranslates Mushkian as Phrygian. Roller (1984: 260, n. 16), slips
when she claims that the name Midas appears in Assyrian texts. Bartl (1994: 514, n. 39)
claims that a Mita of Mushki is mentioned in texts of Shalmaneser III. While I support the
Midas-Mita equation, I realize that on the one hand it is an interpretation, and on the other
it unnecessarily confuses the important issue that the Assyrians never use the term Phrygian
(see also n. 3).
Herodotus (VII.73) and Strabo (VII.3.2) preserved an old Iron Age tradi-
tion that either before or just after the Trojan war the Phrygians migrated
into Anatolia from Thrace or Macedonia, where originally they were named
Brygians. The fact that the Phrygian language is related to the Greek/Balkan
dialects supports the historical tradition (Haas 1970: 33, 5859, 68; Neumann
1988; Drews 1993a: 9). But for the Mushku/i and their name there is no pre-
served tradition, only the Assyrian texts, which do not discuss earlier history
or traditions.
One modern hypothesis holds that the Mushki were related to the Bry-
gians, each constituting one of the many migrating tribes noted in ancient
sources, but that the former had migrated further east (Krte 1904: 910;
Akurgal 1983: 7274; Diakonoff and Neroznak 1985: xixiii). Other scholars
see the Mushki as a distinct people, living in and deriving from the east (e.g.,
Mellink 1965; Laminger-Pascher 1989: 1624). Notwithstanding these con-
flicting views, no scholar has been able to know the actual historical reality.
Perhaps there was a western B/ Phrygian conquest, or a peaceful assimila-
tion, of the eastern Mushki, whether the latter were related or not. Whatever
the historical background, however, if the late eighth century bc equating of
the names makes sense and is accepted, then at least by that time there was
but one polity, recognized by themselves and others as sharing ethnicity, a
common king, an ideology, even a common official languagemanifested
by the wide distribution of texts in the Phrygian language. Whether or
not the Mushki originally spoke a different language is a question directly
related to that concerning their relationship to the Brygians.3
Up to the 1960s no site in central Anatolia, west, east, or south of the
Sangarios and Halys Rivers, was recognized to have yielded material of
any form that could be situated in the time between the Late Bronze Age
destructions and the eighth century bc (Muscarella 1988: 417420). This
picture has been elegantly modified by archaeological research at Gordion.
Gordion is the one site in Anatolia, to date, where unambiguous information
bearing on cultural development within the Dark Age has been revealed.
3 Precisely what the Mushki/Phrygian appellations signify is a fundamental problem
encountered in any discussion of Phrygian history. Perhaps relevant is the fact that the Assyr-
ians, and modern historians, use the term Urartu for the polity identified by its inhabitants
as Biainili (Zimansky 1985: 51, 78). We do not yet know what the Phrygians called themselves,
but it certainly could have been something like B/Phrygians. Diakonoff and Neroznak (1985:
xii) deny this; but Lubotsky (1988: 1314, 16) and Woudhuizen (1993: 6) interpret the word
vrekun in the Areyastis inscription at Midas City to read either (Lubotsky) of the Phrygians,
or (Woudhuizen) the Phrygian, in either case an ethnicon; the Greek Phrugioi would then
be a reflex of this word. This is ingenious and warrants more comment from other linguists.
From a deep test trench cut in 1950 hand-made (HM), coarse dark-ware
sherds were recovered below the eighth century, Early Phrygian, level that
was destroyed ca. 700 bc. It was not, however, until 1960 (and even then ten-
tatively: Mellink 1960: 251; Bittel 1983: 3839; Sams 1992: 5859)4 that the
significance of this ware was first recognized, namely that it was related
to pottery well known from Levels VIIb1 and 2 at Troy and also from the
Balkans, and that it documented the presence of a people living in central
Anatolia after the Late Bronze Age destructions and before the eighth cen-
tury bc.
In 1988, after a 14-year hiatus, field excavations recommenced at Gordion
under the direction of M. Voigt (see Voigt 1994 for the excavation results
utilized hereand bibliography of the 19881990 campaigns). Because of
their significance for the purposes of this article, a summary of the finds
published to date is necessary.
Excavations below the destroyed Early Phrygian settlement revealed sev-
eral semisubterranean structures above the Late Bronze Age surface. They
were furnished with hearths, ovens, bins, and storage pits; this level was
labeled Early Iron Age, Phase 7B. HM pottery was retrieved in situ in and
around the structures, which seem to have been abandoned at different
times. This assemblage of architecture and pottery indicates, to date, the
only known HM culture settlement in Anatolia aside from Troy (in recent
campaigns at Kaman Kalehyk and Bogazky HM sherds have been
recoveredpersonal examinationbut their contexts have yet to be ana-
lyzed; Yakar 1993: 8).
Directly above Phase 7B the investigators recovered a single structure
that utilized different building techniques and internal appointments from
those below; this level is called Early Iron Age, Phase 7A. The structure was
destroyed by a fire and the collapse sealed in a deposit of pottery, which
again represents a new ware, a buff fabric partly wheel-, partly hand-made.
It does not resemble and has no connection with the pottery of the previous
7B phase and, in fact, marks the first appearance of the typical shapes in
use throughout the Early Phrygian settlement down to ca. 700 bc. Above this
structure and cutting into it was another (partly excavated) structure and a
pit, each of which contained both HM and the buff wares, the former surely
indicating (as the burned house did not) that the Phase 7B HM continued
to be manufactured into the next phase.
4 Thus Young (1966: 276277) in his report of a sounding to the Late Bronze Age level
mentions HM wares but does not discuss their significance. See Sams (1988: 910).
Following these two distinct cultural assemblages began a series of strat-

ified building phases collectively labeled Early Phrygian, Phase 6B. The first
stages of Phase 6B initiate the building activity that eventually developed
into the final settlement destroyed ca. 700bc (the latter referred to as Phase
6A).
We turn now to the crucial issue of chronology. It is apparent that at
Gordion itself there is at present no objective criterion for a close dating
of Phase 7B. To obtain a reasonable estimation of absolute dates for this
phase requires looking to Troy, a site that has yielded HM remains within
a stratified sequence. Although it has long been recognized that the HM
at Troy represents an intrusion from the Balkans that followed upon the
destruction of Level VIIa; the dating of this intrusion often has been debated.
In recent years, however, reexamination of the pottery recovered in the
crucial levels has narrowed the area for chronological maneuvering, for it
has been revealed that Mycenaean IIIC pottery was present in both the
destroyed Level VIIa and the following Level VIIb1, that sub-Mycenaean
wares were present in Level VIIb2, and that Greek Protogeometric and
Geometric wares existed in Level VIII (Bloedow 1988: 3233, 3536; Hertel
1991: 139140). This information suggests that Troy Level VIIb1 existed in
the late 12th century bc, and Level VIIb2 existed for some time in the 11th
century bc. Also, the occupation gap alleged to exist between Levels VIIb2
and VIII is either drastically narrowed or eliminated.
It should be possible to invoke the Trojan chronology as a clue for rec-
ognizing the date of the HM presence at Gordion, and thus to posit that
the Phase 7B settlement there was probably founded sometime in the late
12th or the 11th century bc. Sams, however, (1994: 2021) recognized a prob-
lem that relates to the acceptance of exact chronological equations of the
Trojan and Gordion HM wares. Although this ware at Gordion is certainly
related to that recovered in Troy and the Balkans, and all three areas share
certain features, there are more differences between the wares at Gordion
and Troy than between those at Gordion and in the Balkans. To Sams, this
situation could reflect either chronological or contemporary geographical
differences; but if the former, then the Gordion HM ware would probably
have been manufactured later than that at Troy. Nevertheless, it is argued
here that subject to future refinements, the Trojan chronology is relevant for
Gordion and that it may viably function as supporting a working hypothesis
that Phase 7B at Gordion did not come into existence before about 1100 bc,
that is, 100 years after the termination of the Late Bronze Age.
Elegant excavation by Voigt has revealed a post-Late Bronze Age settle-
ment that has narrowed the Dark Age, noncontinuity, gap to 100 yearsa
situation that forces a major modification of the 400 years gap in occupa-
tion previously postulated by Akurgal and supported by myself and others
(e.g., Akurgal 1983; Muscarella 1988: 417420 with bibliography; Mellink 1991:
621).
So remarkable is the coincidence of the archaeological evidence (here
primarily pottery) with the preserved historical tradition that one is moved
to recognize in the Early Iron Age Phase 7B assemblage a precious docu-
mentation of the arrivalI think migration is the relevant wordof the
B/Phrygians from lands in the west (a view approached but not grasped by
DeVries 1987: 7, and Sams 1988: 9, 13; 1992: 59; 1994: 21). That the B/Phrygians
migrated from Europe has been accepted by most scholars (e.g., Krte 1904:
2; Mellink 1965: 317; 1991: 621; Haas 1970: 31, 36; Diakonoff and Neroznak 1985:
xi; Sams 1988: 10; 1992: 59; Laminger-Pascher 1989: 12, 1516, 25, n. 27, 32, 34
36but to her not before the ninth and eighth centuries bc; Yakar 1993:
810, n. 9); a vigorousbut more asserted than demonstratedrejection
of the European migration is given by Drews (1993a; 1993b: 6566, 69);5 Car-
rington (1977: 117) considers the issue open to doubt. The material recov-
ered by archaeology considerably reinforces the historical reality of the
migration tradition: that Phase 7B documents the arrival of the B/phrygians
is probable, surely it is feasible.
5 Drews, of course, had no access to the latest archaeological research of M. Voigt but
in a recent publication (1993b: 65) he ignores the implications of the HM ware already
known from Gordion, which he cites only to dismiss; he ignored this information in 1993a.
Drews seems to conflate two separate matters, the alleged Phrygian migration itself and the
Phrygians alleged role in the destruction of the Hittites. To him, to reject the latter is to
reject the former. A major thrust of Drews 1993b is to deny the role of migrating people in
the catastrophe of ca. 1200bc, and he challenges most ancient references to such events.
Yet, casually and without discussion he automatically accepts Iron Age Greek migrations
(but never using the M-word) to Cyprus and Ionia, Etruscans to Lemnos, and Cretans to
Palestine, calling them colonists and refugeeslabels denied to the Brygians (Drews
1993b: 59, 63, 6869). The idea of a Phrygian migration has certainly not been generally
abandoned by Anatolian archaeologists (Drews 1993b: 65), as his own note 58 should have
indicated.
Relatively recent research is turning up evidence (pottery, statuettes, metal forms) that
indicate a period of strong contacts existed (whatever the causes) between the Balkans and
north and central Anatolia in the Chalcolithic period, from the late fifth through the fourth
millennium bc, probably in this case via the Black Sea. For bibliography see Thissen (1993:
207237); several articles in the special issue on Anatolia and the Balkans in Anatolica
19 (Roodenberg 1993); and S. Steadman in this issue of BASOR (pp. 1332). Because of the
millennia separating these contacts from the Brygian migration, it is not productive to
postulate at this time a long traditional mutual knowledge of the two areas. Yet, one is
intrigued by the events; future research may turn up more Balkan-Anatolian contacts.
An internal clue for the dating of the next period, Phase 7A, also lies in
the pottery. The shapes closely resemble those from late eighth century bc
contexts at Gordion, an indication, perhaps, that not too much time had
elapsed between the floruit of Phase 7A and the late eighth century bc.
If this suggested late dating survives future reviews, one would have to
consider the possibility that the mound had experienced a second gap (an
abandonment) in occupation after the 7B phase. I would estimatethere
is no other method at presentthat not many decades elapsed between
the beginning of Phase 7A and that of Phase 6B; an initial date of ca. 850
bc (at the earliest) for Phase 7A, and a time around 825/800 bc for the
initial Phase 6B construction cannot be too far from the historical reality.
This allows for a century or more (100125 years: not so short a time as some
scholars would perceive) for the full development of Phase 6B, the major,
so-called Early Phrygian, period.6
This is a brief outline of the archaeological evidence known so far from
Anatolia. But other evidence for this period also derives from ancient histo-
rians.
Arrian (II.3) and Justin (XI.7, 35), among others (Roller 1984: 256260,
and n. 4, for a summary of the sources), have preserved a tradition concern-
ing the establishment at Gordion of Phrygian kingship, i.e., the formation of
a unified Phrygian state, sometime in the Dark Age, within the period under
review. The tradition, as preserved, records that strife broke out among the
Phrygians and was resolved by a poor farmer, either Gordios, or his son
Midasthe sources disagreewho by arriving at a propitious moment at
Gordion in a cart or wagon drawn by an ox fulfilled a prophecy that conse-
quently led to his election as the king of Phrygia. In some sense this tradition
must surely reflect remembrance of an historic event.7
6 It would seem that the term Early Phrygian used to designate Phase 6B could be
misleading; Voigt uses it to refer to the historical period. The chronology presented here is
not accepted by the excavator of Gordion. Voigt (1994: 270) allows considerably more time for
Phases 7A and early 6B: she begins Phase 7A in the early tenth century bc, with no gap, and
she begins Phase 6B ca. 900/850bc. Yaker (1993: 11) believes that the 7A phase may represent
the arrival of the Phrygians in the late second, early first millennia bc; he paradoxically says
little of Phase 7B except that it represents Thracian 12th century bc elements and is pre-
Phrygian (Yakar 1993: 10, n. 10), which (in the context, confusingly) omits discussion of the
implications of HM and its possible Brygian connection.
7 Roller (1984: 263270) presents a sununary of possible religious and folklore elements
preserved behind the scenes in the story, the relationship of Gordion to Gordios, the historic-
ity of Gordios (a fiction to her), the significance of Kybele, and so forth. None of this need be
discussed here except to note that the basic core element of the traditionstate formation
From the meager evidence at hand, it is possible to suggest the following

scenario: the Phrygians were a self-recognized ethnic group not ruled by a
central authority, a king; civil strife developed; to resolve the resulting prob-
lems a conscious act of social restructuring and societal transformation was
implemented. Manifest in this deed was the elimination of former institu-
tions of authority, recognized to be weak, and the subsequent subordination
of the society to one central, absolute authorityin perpetuity. The transac-
tion must have brought about, among other things, more internal security,
protection from disruptive external forces, and a more stable mechanism
for administrative and information dissemination procedures.
We know and can infer not much else. Appropriate questions that define
the problems and assess the available information surface, but the answers
come out either as very tentative or as more questions.
Some of the questions that arise from the tradition as preserved include:
was the Ur-king an indigenous Phrygian or a member of a newly arrived
related group or tribe? The arrival of the future king in a wagon might
suggest the latter. Both philologists and archaeologists independently posit
more than one European incursion into Phrygia over time (Haas 1970: 31,
36 [he sees the second migration as occurring in the ninth century bc];
Carrington 1977: 118119; Sams 1988: 1213; 1992: 59; cf. Laminger-Pascher
1989: 32, 3435, who argues for one, ninth century bc, migration). Is the
crucial wagon motif connected to a migration that introduced new forces?
Roller (1984: 262, n. 29) mentions a second-century bc Macedonian historian
who recorded that Midas migrated to Anatolia in the famous wagon; but,
as she notes, this could be a much later attempt to support a parallel with
Alexanders travels. Because in Arrians version the wagon is associated with
an eagle, the two also have been interpreted to represent a close connection
of Midas with a deity and his chariot (e.g., Krte 1904: 16; Roller 1984: 266
269). If this be the case, the connection of a first or a new king with a deity
does not surprise us. Midas was associated in Greek legend with several
deities (Eitrem 1932: 15311534) and probably, like other Near Eastern and
Greek kings, held both religious and secular powers that were not always
separated (Runciman 1982: 357, 362).
There remains an interesting and intriguing historical parallel to the
wagon element in the Phrygian tradition. Calmeyer (1974: 63, n. 57) cogently
raised the possibility that the Phrygian became king with the help of his
from a nonstate societycannot be readily dismissed as fiction of mere folktale, as earlier

writers realized, e.g. Krte 1904: 1315.
wagon in the same manner as did Rusa I, king of Urartu.8 In Sargon IIs
report to Assur on his eighth campaign in 714bc he records his capture of the
Urartian holy center of Musasir. Taken as booty, among many other items,
was a statue group of Rusa, with two of his horses, (and) his charioteer,
which was inscribed With my two horses and one charioteer, my hand
attained the kingdom of Urartu (Luckenbill 19261927: II, 173). What the
attainment entailed is not recorded; but just the year before, in 715bc, Rusa
survived a brutal revolt (Lanfranchi 1983) by escaping into the mountains,
and it was probably that event that was commemorated with the creation
of the statue group.9
What stands out is that the Phrygian king could have claimed that with
my one ox and one charioteer, my hand attained the Kingdom of Phrygia,
for that meaning is precisely what is preserved in the tradition. The histor-
ical parallel is certainly real, for a wagon and its connection to kingship
are associated with turmoil and a favorable resolution in both the Phry-
gian and Urartian events, although the specifics are different. In one case a
commoner who peacefully and innocently enters a city is made king, in the
other a king saves his reign by escaping, exiting, to the mountains. We are
dealing here with two different and unconnected incidents, one an histori-
cal report of a contemporary occurrence, the other preserved through local
tradition and perhaps not revealing the full circumstances that the wagon
episode entailed. Nevertheless, the possibility of condensation in the tradi-
tion does not compel a denial that an historical reality involving a wagon and
its occupant occurred. It would not be fruitful, though, to speculate about
the incidents that might lie behind the Phrygian episode based on the his-
torical Urartian experience.
8 Calmeyer (1974: 63, n. 57) also cites as a parallel Herodotus (III: 8586) who reports
that Darius attained kingship with a ruse involving a horse and a groom. Here there is only
a horse, not a wagon, but there is an essential formal similarity to the Gordion and Rusa
events. Calmeyer (1974: 6364) also links the Gordion wagon and its eagle (from Arrian) with
a deity and notes the wagon symbolizing the victory-bringing power of the highest deity.
Surprisingly, he does not pick up Krtes (1904: 16) firm claim that the Phrygian wagon is
Ein genaues Gegen stck to the leere Wagen des persischen Himmelsgottes Krte also
believed that the wagon was originally associated with Zeus and later associated with the
farmer king.
9 We are informed that to save his life Rusa escaped alone, but surely he had his chariot
and charioteer. Rusa suppressed the revolt and lived to commission the statue group and
later fight Sargon II. Lanfranchi does not mention the Musasir statue but Barnett (1982: 351)
mentions it in connection with the revolt. The statue group would then have been less than
one year old when taken to Assyria by Sargon.
What can one say about the nature of the prekingship societyfor exam-
ple, what authority existed? were there tribal or clan chiefs at Gordion? was
Gordion the residence of a paramount chief? While it may satisfy neoevo-
lutionary theory to add still another chief or chiefdom to the many per-
ceived in anthropological literature, it may not be historically accurate to
do so.10 A neoevolutionary model would posit that the Phrygians should
have experienced first tribal and then chiefdom stages of government before
becoming a state, i.e., developing into a kingdom. Some anthropologists
reject this deterministic, integrative position and recognize the difficul-
ties in ascertaining not only what a chiefdom actually is, and the man-
ner in which it allegedly becomes, or is different from, a state, but also
that more than one state formation model exists. Ethnology can provide
analogies but not equations; the Phrygians were Thracians, not Polyne-
sians.
The act of election as a process of societal transformation need not be
an impediment to accepting the legends historical core. Aristotle (Politics:
1285: 3,7) discusses the existence both of elective, nonhereditary dictators
and elected, hereditary kings in Greek city-state formation (see also Runci-
man 1982: 357). In these instances the elected king had previously performed
some crucial services for his society. The tradition does not define the nature
of the political strife at Gordion, but the society was fundamentally dis-
rupted. Anthropologists suggest that a number of dramatic causes can result
in state formation, among them population growth, class stress, technolog-
ical or subsistence pressures, and external pressures or threats of warfare.
Which of those events (if any) may have caused the strife? All we can dis-
cern is that an individuala peasant, not a noblemancame to the aid of
his fellow Phrygians. We can also infer that he performed some significant
acteven fulfilling an oracle.
Nor does the tradition inform us about the role of the other Phrygian cen-
ters. If we are justified in using the HM ware as the criterion for determining
the presence of Brygians at other centers, we again ask questions that can-
not be answered: did the Phrygians at Bogazky and Kaman Kalehyk
neither site geographically close to Gordionand other centers still unrec-
ognized play a role in the formation of the Phrygian state? Was the strife
10 A good picture of the nature of pre- and poststate societies may be found in a number
of sources, notably the publications of R. Carniero, T. Earle, M. Fried, G. Johnson, P. Kohl,

L. Rouse, F. Service, and N. Yoffee. For a comprehensive bibliography of this material, see
Yoffee (1993: 7478).
limited to Gordion or were the other centers involved? How and when did
those centers accept the kingship organized at Gordion?
What information do the archaeological data yield regarding these prob-
lems? Can a pre- and postkingship pattern be observed in the excavated
remains at Gordion?
It is probable that the population settled in the initial Early Iron Age
phase, 7B, did not have a king. The material remains show that Phase 7B
was surely a village, as Henrickson (1994: 108) has posited that the pottery
was made by individual households. Hierarchical architecture has not been
exposed and no burials exist to reveal the mortuary furniture of an elite. No
archaeologist would argue from the remains that kingship existed in Phase
7B (the peasantlike character of the prekingship Phrygians in the preserved
tradition was noted by [Krte 1904: 1516]).
The one structure fully excavated in Phase 7A was burned, and the charac-
teristic buff ware is interpreted (again by Henrickson [1994: 109110]) to have
been more complex in production (mass produced?). Here, too, there are no
clues in the architecture or in burial customs to suggest conclusions about
the nature of the society. Whether the burning of the Phase 7A structure
reflects the strife that is noted in the tradition must remain an intriguing,
but yet another unanswered, question. Further excavation will surely pro-
vide information about the extent of the burning of that phase. But what is
significant about the existence of the HM and buff ware recovered together
in the later part of Phase 7A (above) is that some element of cultural conti-
nuity from the earliest settlement is revealed.
The first unambiguous, archaeologically documented indication of orga-
nized political and social activity in Gordions history was the initiation of
the massive building program around 825800bc. The Phase 6B master plan
began tentatively, but within a short time represents a dynamic conception
absent in the earlier periods and it reflects the existence of an authority
commanding power and concomitant wealth, and an urban concept. In its
developed stage one encounters monumental architecture; massive fortifi-
cation walls (completely rebuilt at least once); royal burials (see below); the
presence of a writing system; specialists such as architects and engineers,
many artisans working in metal, ceramics, ivory, stone, textiles, and wood
carving; and also the availability of objects from foreign centers, probably
gift exchanges. It tentatively follows, then, that kingship could have arisen
perhaps late in Phase 7A or at the beginning of Phase 6B. If the beginning of
the 6B phase is indeed the actual first stage of the master plan, the evidence
could be interpreted to indicate late Phase 7A as the time when kingship
arose. In absolute dates this would be approximately 825 bc.
One hypothesis about Phrygian state formation that comes easily to mind
will not cause much opposition: the political transformation experienced
by Gordion-Phrygia would not have been that of primary state formation,
an indigenous, self-generated, and independent social evolution, but of
secondary state formation, in which state origins are attributed to pressures
or models derived from previously existing states. Such a development could
occur quickly and need not be gradual (Runciman 1982: 356). In the mid-
and late ninth century bc, kingships existed in neighboring areas, at least
in Assyria, North Syria, and Urartu, and perhaps also in Tabal. Aside from
Assyria, with which Phrygia (Mushki) had hostile relations (and perhaps
also normal contacts), there is no information concerning the intensity of
interaction between Phrygia and the other polities in the ninth century bc.
But surely their existence could not have remained unknown; one or all of
these states could have been a model for those who chose a citizen to be
their first king.
A potentially important historical implication is suggested by the chro-
nology advanced here for the advent of the Phrygian state (ca. 825 bc). The
dates correspond either to the last years of Shalmaneser III (858824bc) or
to the reign of Shamshi-Adad V (823811bc). For some time it has been rec-
ognized that during the reign of Shalmaneser III the Urartians consolidated
themselves and seem for the first time to have achieved a unified state, ruled
by one king (Zimansky 1985: 4850). This event occurred at the very time
proposed here for the Phrygians consolidation under one king. Two sec-
ondary state formations would then have occurred in Anatolia in the second
half of the ninth century bc. More refinement on chronology is necessary;
but even if a higher chronology is indicated at Gordion, the proximity in time
of the two events will not be essentially altered. The populations of central
and northeastern Anatolia were in continuous conflict with Assyria, a ten-
sion that may have provoked (in part) parallel social transformation in each
area in the second half of the ninth century bc (for Assyrian aggression and
Urartu see Zimansky 1985: 4850).
The first Phrygian king or his son (perhaps either the great-grandfather
or grandfather of the Midas who perished in the destruction of ca. 700bc)
or both should be considered as the mind(s) that conceived and initiated
the construction of the Phase 6B settlement. This major project could not
have been conceived and carried out in a noncentralized society. The early
modern historians of the Phrygians, Krte (1904: 1314, 21), Swoboda (1912:
15901591), and Eitrem (1932: 15361538) interpreted the ancient sources to
indicate that the historical Midas (ruled ca. 738696 bc) was in fact the
second of that name, the first having been mentioned in the traditional
story. They also believed, correctly, I think, that it makes sense to accept
Justins version that a Gordios, not a Midas, was that first king.11
The earliest royal burial (so determined by its size, second only to MM,
and its rich contents) so far recognized at Gordion is that in Tumulus W,
plausibly dated to the years around 750bc12 By this time some of the Phase
6B settlements structures were functioning units. Positing a few decades
for a reign, the beginning date for the buried kings rule would have been
ca. 780 bc. Since many tumuli remain to be investigated, it cannot be
argued on archaeological grounds that the tomb contained Phrygias first
kingwe do not yet know how and when this form of burial first came to be
used at Gordion. However, based on the chronology posited here, the man
buried in Tumulus W could have been either the second or third king to
reign at Gordion: Midas I or Gordios II.13
The historical importance of Tumulus Ws chronology (whatever ones
opinion of that date is) is significant for two reasons: its proposed date
(750bc) is a post quem non for the existence of kingship in Phrygia, and the
estimated date (780bc) for the buried kings first regnal year is a viable, if
hypothesized, earlier post quem non.
11 I had reached this view before going back and rereading these German scholars. It fol-
lows that I reject Rollers opinions (1984: 260, 263) that Gordios was a fictitious name, e.g.,
Roller also posits that the first Phrygian king would have had a Balkan, not an Anatolian name
(Midas), a view that ignores evidence of the rather long sojourn of the Phrygians in Anato-
lia: the new excavation evidence interpreted here further argues against the need to seek a
Balkan name (see n. 3, above). However, I am informed that some linguists believe that Gor-
dion derives from gorod or gradwhich if correct, signifies that the name Gordion/Gordios
is indeed Balkan.
12 Independently dated this late by both my research (Muscarella 1982: 8): 750/740 bc, and
that of R.M. Boehmer (1984: 259260): ca. 740bc. Of the many excavated tumuli at Gordion,
about 11 date to the predestruction periods: G, P, Q, S, MM, W, Y, KIII, KIV; X and KY (not yet
fully published) have also been dated to this period by DeVries (1977: 9, n. 11), and Sams (1994:
192196), although each has a different list (see also Muscarella 1967: 111).
13 This raises again the question of who was buried in Tumulus MM. I have argued
elsewhere (Muscarella 1982: 910), and still hold the view, that it could not be Midas because
he was alive when the burial was arranged, sometime between 720 and 700 bc. It follows
chronologically that it was he who must have ordered the tomb and tumulus to be built. Other
scholars, however. date MM to a time just after the destruction, and consequently consider it
to be Midas burial (e.g., DeVries 1977: 9; Mellink 1991: 634; Hawkins 1994: 273; see the judicious
comments of Sams 1994: 196). The only way for archaeologists to agree that MM contained the
burial of Gordios II (who presumably died before his son began to reign) is for its deposition
date to be raised to ca. 740bc at the latestbut which on the basis of present knowledge will
not satisfy any archaeologist. If neither Midas nor seemingly his father, Gordios II, is buried
in MM, the question of who is interred there remains intriguing. The burial was manifestly
arranged for a revered person who died in his 60s, a man Midas wished to honor by erecting
the largest tumulus hitherto known in Anatolia.
In the chronology suggested here, no change is required in the sequence

proposed by the early modern historians for the historical king, i.e., Midas II.
If this Midas ruled from ca. 738 to 696bc, on a conventional 30 years to a
generation estimation, his father will have reigned from ca. 770 to 738bc,
his father from ca. 800 to 770bc, and his father from ca. 830 to 800bc.
Even adding five to ten years to each reign (since Midas II reigned for more
than 40 years) brings us no earlier than the 850s bc. This chronology allows
three kings to precede the historical king: his father, grandfather, and great-
grandfather, or (following the Phrygian tradition of alternating the royal
names) Gordios I, Midas I, and Gordios II.
Obviously, scholars who accept a longer chronology between the histori-
cal Midas and the first king, or who begin the dynasty with a Midas, will add
one or more generations and adjust the reign (Roman) number to be placed
after the historical Midas. I am aware that the sequence advanced here is
shorter than the number of kings who reigned in Urartu and Assyria during
the same time span proposed, eight for the latter, six for the former. Thus,
while at present suggesting four generations, I surely cannot exclude the
possible existence of another Gordios-Midas. This would make the historical
king Midas III, not Midas II. Recent archaeological research has demon-
strated that the Phrygian kingdom had a longer life than some scholars had
previously argued, but the fact remains that the predestruction kingdom
experienced, as Akurgal neatly expressed it, a glorieuse mais trs courte vie
(Muscarella 1988: 420).
Although only a small body of information is currently available on the
Iron/Dark Age, viable information and hypotheses have been teased from
the data. It has been demonstrated that students of Anatolian archaeology
and history know far more about the Dark Age now than was known only a
decade ago. Excavations are making the Dark Age brighter.14
References
Akurgal, E.
1983 Das Dunkle Zeitalter Kleinasiens. Pp. 6778 in Griechenland, die gis und
die Levante whrend der Dark Ages vom 12. bis zum 9. Jh. v. Ch., ed. S. Deger-
Jalkotzy. Vienna: sterreichische Akademie der Wissenschaften.
14 I thank Kenneth Sams, Elizabeth Simpson, and Mary Voigt for their cooperation in
sharing their ideasnot always in agreement with mineand for helping to make this a
better article than it would have been without their concern.
Astour, M.
1965 New Evidence on the Last Days of Ugarit. American Journal of Archaeology
69: 253258.
Barnett, R.D.
1982 Urartu. Pp. 314371 in The Cambridge Ancient History, vol. 3, part 1, eds.
I. Boardman, I.E.S. Edwards, N.G.L. Hammond, and E. Sollberger. 2nd. ed.
New York: Cambridge University.
Bartl, K.
1994 Die Frhe Eisenzeit in Ostanatolien und ihre Verbindungen zu den Benach-
barten Regionen. Baghdader Mitteilungen 25: 473518.
Bittel, K.
19761977 Das Ende des Hethiterreiches aufgrund archo-Zeugnisse. Jahresbe-
richt des Instituts fr Vorgeschichte der Univ. Frankfrt a.M.: 3656.
1983 Das archologische Situation in Kleinasien um 1200 v. Chr. und whrend der
nachfolgenden vier Jahrhunderte. Pp. 2565 in Griechenland, die gis und
die Levante whrend der Dark Ages vom 12. bis zum 9. Jh. v. Ch., ed. S. Deger-
Jalkotzy. Vienna: sterreichische Akademie der Wissenschaften.
Bloedow, E.F.
1988 The Trojan War and Late Helladic III C. Prhistorische Zeitschrift 63: 2352.
Boehmer, R.M.
1984 Uruk-Warka XXXVI: Survey des Stadtgebietes von Uruk. Baghdader Mit-
teilungen 15: 113140.
Calmeyer, P.
1974 Zur Genese altiranischer Motive II. Der leere Wagen. Archaeologische Mit-
Carrington, P.
1977 The Heroic Age of Phrygia in Ancient Literature and Art. Anatolian Studies
27: 117126.
Caubet, A.
1992 Reoccupation of the Syrian Coast after the Destruction of the Crisis Years.
Pp. 123131 in The Crisis Years: The 12th Century B.C., eds. W.A. Ward and
M. Joukowsky. Dubuque, IA: Kendall/Hunt.
Deger-Jalkotzy, S., ed.

1983 Griechenland, die gis und die Levante whrend der Dark Ages vom 12. bis
zum 9. Jh. v. Chr. Vienna: sterreichische Akademie der Wissenschaften.
DeVries, K.
1987 Phrygian Gordion Before Midas. Pp. 612 in Anatolian Iron Ages, ed. A.
ilingiroglu. Izmir: Ege Universitesi Edebiyat Fakultesi.
Diakonoff, M.I., and Neroznak, V.P.

1985 Phrygian. Delmar, NY: Caravan.
Drews, R.
1993a Myths of Midas and the Phrygian Migration from Europe. Klio 75: 926.
1993b The End of the Bronze Age: Changes in Warfare and the Catastrophe ca. 1200
B.C.. Princeton: Princeton University.
Eitrem, S.
1932 Midas. Paulys Real-Encyclopaedia 15: 15261540.
Freu, J.
1988 La Tablette RS 86.2230 et la phase finale du Toyaume dUgarit. Syria 65: 395
398.
Gunter, A.
1991 The Bronze Age: The Gordion Excavations Final Reports III. University Muse-
um Monograph 73. Philadelphia: University Museum, University of Penn-
sylvania.
Haas, O.
1970 Das Problem der Herkunft der Phryger und ihrer Beziehungen zu den
Balkanvlkern. Acta Antiqua Academiae Hungaricae 18: 3169.
Hawkins, D.
1994 Mita. Reallexikon der Assyriologie 8, 3/4: 271273.
Hawkins, J.D.
1982 The Neo-Hittite States in Syria and Anatolia. Pp. 372441 in The Cam-
bridge Ancient History, vol. 3, part 1, eds. J. Boardman, I.E.S. Edwards,
N.G.L. Hammond, and E. Sollberger. 2nd ed. New York: Cambridge Univer-
sity.
Henrickson, R.
1994 Continuity and Discontinuity in the Ceramic Tradition of Gordion during
the Iron Age. Pp. 95129 in Anatolian Iron Ages 3, eds. A. ilingiroglu and
D.H. French. Ankara: British Institute of Archaeology.
Hertel, D.
1991 Schliemanns These vom Fortleben Troias in den Dark Ages im Lichte
neuer Forschungsergebnisse. Studia Troica I: 131144.
Hoffner, H. A, Jr.
1992 The Last Days of Khattusha. Pp. 4652 in The Crisis Years: The 12th Century
B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/Hunt.
Korte, A.
1904 Geschichte Phrygiens. Pp. 127 in Gordion: Ergebnisse der Ausgrabung im
Jahre 1900, eds. G. and A. Krte. Jahrbuch des Kaiserlich Deutschen Archol-
ogischen Institts, Ergnzungsheft 5. Berlin: Reimer.
Laminger-Pascher, G.
1989 Lykaonien und die Phryger. Vienna: sterreichische Akademie der Wis-
senschaften.
Lanfranchi, G.B.
1983 Some New Texts about a Revolt Against the Urartian King Rusa I. Oriens
Antiquus 22: 123136.
Lubotsky, A.
1988 The Old Phrygian Areyastis-Inscription. Kadmos 27: 926.
Luckenbill, D.D.
19261927 Ancient Records of Assyria and Babylonia III. 2 vols. Chicago: Univer-
sity of Chicago.
Mellink, M.
1960 Review of C. Blegen et al., Troy IV. Bibliotheca Orientalis 17: 249253.
1965 Mita, Mushki and Phrygians. Anadolu Arastirmalari 2: 317325.
1991 The Native Kingdoms of Anatolia. Pp. 619643 in The Cambridge Ancient
History, vol. 3, part 2, eds. J. Boardman, I.E.S. Edwards, N.G.L. Hammond,
and E. Sollberger. 2nd ed. New York: Cambridge University.
Muscarella, O.W.
1967 Phrygian Fibulae from Gordion. Monograph Series (Colt Archaeological
Institute) 4. London: Quaritch.
1982 King Midas Tumulus at Gordion. Quarterly Review of Archaeology 3, no. 4:
710.
1988 Bronze and Iron. New York: Metropolitan Museum of Art.
1989 King Midas of Phrygia and the Greeks. Pp. 333344 in Anatolia and the
Ancient Near East, eds. K. Emre, M. Mellink, B. Hrouda, and N. zg,
Ankara: Trk Tarih Kurumu.
Neumann, G.
1988 Phrygisch und Griechisch. Vienna: sterreichische Akademie der Wissen-
schaften.
Otten, H.
19761977 Zum Ende des Hethiterreiches aufgrund der Bogazky-Texte. Jahres-
bericht des Instituts fr Vorgeschichte der Univ. Frankfurt a.M.: 2235.
1983 Die letzte Phase des Hethitischen Grossreiches nach den Texten. Pp. 1324
in Griechenland. die gis und die Levante whrend der Dark Ages vom 12.
bis zum 9. Jh. v. Ch., ed. S. Deger-Jalkotzy, Vienna: sterreichische Akademie
der Wissenschaften.
Postgate, N.
1973 Assyrian Texts and Fragments. Iraq 35: 1336.
Roller, L.
1984 Midas and the Gordion Knot. Classical Antiquity 3: 256271.
Roodenberg, J., ed.

1993 Anatolia and the Balkans: Symposium on Pre-Bronze Age Relations held in
Istanbul, 1822 November 1991. Anatolica 19.
Runciman, W.G.
1982 Origins of States: The Case of Archaic Greece. Comparative Studies in Society
and History 24: 351377.
Sams. G.K.
1988 The Early Phrygian Period at Gordion: Toward a Cultural Identity. Source 7,
no. 3/4: 915.
1992 Observations on Western Anatolia. Pp. 5660 in The Crisis Years: The
12th Century B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/
Hunt.
1994 The Early Phrygian Pottery. The Gordion Excavations, 19501973: Final Re-
ports, 4. 2 vols. University Museum Monograph 79. Philadelphia: University
Museum, University of Pennsylvania.
Swoboda, H.
1912 Gordios. Paulys Real-Encyclopaedie 7: 15901594.
Thissen, L.
1993 New Insights in Balkan-Anatolian Connections in the Late Chalcolithic: Old
Evidence from the Turkish Black Sea Littoral. Anatolian Studies 53: 207
237.
Voigt, M.
1994 Excavations at Gordion 19881989: The Yassihyk Stratigraphic Sequence.
Pp. 265293 in Anatolian Iron Ages 3, eds. A. ilingiroglu and D. French.
Ankara: British Institute of Archaeology.
Ward, W.A., and Joukowsky, M., eds.

1992 The Crisis Years: The 12th Century B.C. Dubuque, IA: Kendall/Hunt.
Woudhuizen, F.C.
1993 Old Phrygian: Some Texts and Relations. Journal of Indo-European Studies
21: 125.
Yakar, J.
1993 Anatolian Civilization Following the Disintegration of the Hittite Empire:
An Archaeological Appraisal. Tel Aviv 20: 328.
Yoffee, N.
1993 Too Many Chiefs? (or, Safe Texts for the 90s). Pp. 6078 in Archaeological
theory: who sets the agenda?, eds. N. Yoffee and A. Sherratt. New York:
Cambridge University.
Yon, M.
1992 The End of the Kingdom of Ugarit. Pp. 111122 in The Crisis Years: The
12th Century B.C., eds. W.A. Ward and M. Joukowsky. Dubuque, IA: Kendall/
Hunt.
Young, R.S.
1966 The Gordion Campaign of 1965. American Journal of Archaeology 70: 267
278.
Zimansky, P.
1985 Ecology and Empire: The Structure of the Urartian State. Studies in Ancient
Oriental Civilization 41. Chicago: Oriental Institute of the University of
Chicago.
chapter eighteen
THE DATE OF THE DESTRUCTION OF

THE EARLY PHRYGIAN PERIOD AT GORDION*
Abstract: For decades the date of the destruction of the Phrygian site of Gordion
in central Anatolia has been accepted to have occurred ca. 700 bc or later. In 2001
the Heidelberg Akademie der Wissenshaften laboratory released the results of
C 14 analyses of material from the destruction level indicating a date of 830 to
800bc. Members of the Gordion excavation team accepted the new chronology.
This high chronology significantly alters the dating of all the artefacts recovered
in the destruction level, affects knowledge of the specific Gordion phase in which
King Midas reigned and the proper assignment of royal tumuli burials to a specific
culture level, and mandates a major evaluation of Anatolian Iron Age history and
archaeology. The present study examines artefacts recovered from the destruction
level and the tumuli and concludes that the destruction occurred at the time
previously recognised. It is argued that the C 14 determination does not reflect the
historical reality.
The Background
In January 2001 the Heidelberg Akademie der Wissenschaften C 14 labora-

tory reported to Mary Voigt the Field Director of the Gordion excavations
that organic samples secured from the Early Phrygian (EP) destruction level
(DL; Fig. 1) at Gordion yielded cal dates of 830800 bc (using 2 sigma dates,
98% probability). This chronology was enthusiastically reported at the Van
Iron Age Symposium in August 2001, then subsequently published as an
historical reality by S.W. Manning et al. and by K. Strobel.1 And Voigt has
unhesitatingly reported both her (and that of other team members) absolute
support of the 9th century DL chronology in a number of public lectures, dis-
cussed it with colleagues, and recently shared it at large in a (flawed) British
television documentary called Midas Revealed.
This C 14-based adjustment has important consequences, for it formu-
lates a chronology that is 100150 years earlier than concluded in all previous
* This article originally appeared as The Date of the Destruction of the Early Phrygian
Period at Gordion, Ancient West and East 2, no. 2 (2003): 225252.

1 Manning et al. 2001, 2534; Strobel 2002, 489.
570 chapter eighteen
Fig. 1. Gordion destruction level (DL) Plan.
chronological discussions by Gordion scholars.2 It changes all suppositions

previously held for many decades about the early history of Phrygia, the time
of its state formation, the specific Gordion level to be associated both with
King Midas and the construction of the tumulus burials, as well as the his-
tory of the Anatolian Iron Age in general. Moreover, of great significance for
Near Eastern archaeology, the new C 14-based destruction date concomi-
tantly requires dismissal of the hitherto recognised chronology of all the
2 See Sams 1979, 46; 1993, 549, 551; DeVries 1990, 388390; Muscarella 1995, 97; Voigt 1994,
272273; 2000, 41, 52, note 11.

destruction of the early phrygian period at gordion 571
artefacts recovered in the DL, which includes Phrygian inscriptions. The

generally accepted date for the DL has been ca. 700bc, or to some either
696 or ca. 676bc (the death of Midas is also thought to be associated with
the destruction of Gordion but this issue will not concern us here).
Presented together with the announcement of the new C 14 results is the
claim that relevant dendrochronological data exist that support, in fact rein-
force this chronology. In the same sentence announcing the new, high C 14
date Manning et al. inform us3 that the last preserved ring from timbers
from Terrace Building 2A in the DL yielded the date 883 + 4/-7 bc, which
date they proclaim approximates construction of this building. The expres-
sion approximates presumably refers to missing terminal rings that would
give a specific cutting date, but which this sample does not have. In any
event in most cases it is impossible to estimate the number of missing rings
from wood without bark, and the published cutting is nothing more than a
post quem indication. There is also a flawed methodology inherent in this
chronological conclusion, for it is a bald non-sequitur from previous state-
ments apropos dating buildings at Gordion from tree rings, whether with
terminal rings or not. P. Kuniholm wrote unambiguously It is clear that
wood was reused at Gordion, sometimes many centuries after it was cut.
[CC3] is the best example of this.4 Three pieces of wood recovered here
were cut about four centuries earlier than other samples in the same struc-
ture; If only the latter had been collected, the result would have been an
entirely erroneous notion about the date of CC3. Similar disparities in cut-
ting dates for wood used in a single structure are documented at Gordion in
the Middle Phrygian (MP) period for at least three other structures, involv-
ing 400500 years difference.5 Put succinctly, the wood in DL structures has
no absolute value either in dating the DL or in supporting a C 14 date, and
should not have been published as a dendrochronological datum in the first
place; it will not concern us further.6
Immediately upon receipt of the Heidelberg report members of the Gor-
dion excavation team unhesitatingly accepted the new chronology and
rejected previously held positions expressed in earlier publications. One
obvious change demanded by the C 14 chronology is the date of the MP
rebuilding of the citadel directly above the DL. The MP citadel was built
3 Manning et al. 2001, 2534.

4 Kuniholm 1988, 8; Kuniholm was a co-author in Manning et al. 2001.
5 Kuniholm 1988, 67, and chart.
6 For a critique of (other) premature announcements by Kuniholms laboratory team re
dendrochronologically determined dates and subsequent retractions, see James 2002, 18.
over a massive clay deposit laid down over the DL and in some areas is
more than 5m deep.7 Voigt had earlier demonstrated that this rebuilding
occurred very soon after the destruction, and not, as had been maintained
by Rodney S. Young (and generally accepted over the years), after a long
period of elapsed time: the MP rebuilding began almost as soon as the ruins
[DL] had cooled 8 and there is little or no gap between the [DL] fire and
the reconstruction of the MP citadel.9 Consequently the MPs construction
is now to be considered as having commenced in the late 9th or early in
the 8th century bc, as pointed out recently by Strobel.10 In this scenario it
was the MP, not the EP, citadel that was the locus from which King Midas
reigned, although he would not have been responsible for the rebuilding,
which would have preceded his reign by six to nine decades. The earliest his-
torical reference to Midas is 717 bc, another, the last known to us, occurred
in 709bc.11 Based on ancient historians, Midas probably began his reign
ca. 740 or a little later, surely before the time of Sargon II. Consequently, the
rebuilder would have been Midas (very young!) father or his grandfather,
possibly even great-grandfather.
While the new C 14 chronology does not alter the 8th century artefact-
based dating recognised for the excavated tumulibased both on den-
drochronology and internal artefact analysis, including imported material
(see below), it does require that these burials are now to be associated with
the middle-late phase of the MP floruit, and not as hitherto accepted with
that of the EP.
Given the unexpected and complex issues introduced by the ad hoc
acceptance of the Heidelberg C 14 analysis, it seems appropriate to (re)ex-
amine the artefacts recovered in the DL specifically with reference to their
archaeologically determined chronologies, that is, independently of the
accepted C 14 date. Acceptance of the new chronology concomitantly de-
mands that all the DL artefacts are to be considered 9th century productions.
This is all the more interesting inasmuch as Gordion excavation team schol-
ars had supported in various publications an 8th century artefact-derived
chronology for these same DL artefacts.12 (Some of the DL artefacts will not
7 Voigt and Young 1999, 208209.

8 Voigt and Young 1999, 202203.
9 Voigt and Henrickson 2000, 52; also DeVries 1998, 2, 5.
10 Strobel 2002, 89.
11 Mellink 1991, 622, 624, 634; Ivantchik 1993, 69.
12 For example, Sams 1994, 1. I gave a preliminary report on the issues discussed here in a
paper delivered at the AOS meeting in Nashville, Tennessee: see AOS Abstracts, April 4th7th
2003, 31.
be discussed here, viz. animal figurines.13 Represented here are at least four
different styles, at least one of which reflects an Aegean background, as Maya
Vassileva has demonstrated in her study of them; forthcoming).
The Fibulae
Fibulae occur at Gordion in DL contexts and in tumulus burials (see Mus-

carella 1967, 1519, 2125, pls. IXV for the specific fibula types mentioned
below; I use the Blinkenberg and Muscarella (1967) classification, Type XII
with subtypes, to define the Phrygian fibula forms. E. Caners 1983 work is of
fundamental importance for Anatolian fibula studies, but his classification
system is cumbersome). The tumuli of the 8th century are the only relevant
ones to be discussed here; they include (in no relative chronological order):
Tumulus III, IV, P, MM, W, G, S and S1. In 1988 a tumulus was excavated at
Mama Creek, 1km south-east of the Gordion citadel mound; it contained 80
Phrygian fibulae.14
All scholars who worked on Gordion material and archaeology agree that
the tumuli were constructed in the second half of the 8th century bc, and
that the earliest tumulus excavated is W, the second largest tumulus at the
site after MM. Young originally dated it to the end of the 9th or early 8th
century bc because he believed that a bowl recovered here was related to a
9th century bc Assyrian vessel.15 This chronology was in general accepted by
M. Mellink, who also concluded the tumulus was built a generation before
Midas; K. DeVries dated W decades if not some generations before the
DL; K. Sams and E. Kohler avoided absolute chronology; and Caner accepts
a first half of the 8th century date;16 I dated W to 750/740bc +/-, a date
independently assigned also by R. Boehmer.17 All Gordion scholars agree that
the Tumulus W fibulae are the earliest examples recognised from the tumuli
(including possibly the earliest fibula type, XII, 7A, known at Gordion).
Not disputed is the recognition that all the other tumuli containing
fibulae (the relevant ones are Tumulus MM, III, IV, P, G, S and S1) were
13 Young 1962, 166, pl. 48, figs. 21, 22, Prayon 1987, Taf. 2125; Sams and Temizsoy 2000,
nos. 58, 59.

14 Temizsoy 1992; 1993.
15 Young 1981, 198199. For discussion of this vessels chronology, see Muscarella 1984, 418.
16 For Mellink, DeVries, Sams and Kohler, see their notes in Young 1981, 270, 272, 198199,
215, 239; also Sams 1994, 17, 19; Caner 1983, 6, 54.
17 Muscarella 1982, 9; 1986, 195; 1988a, 427, note 5; 1988b, 182183; 1995, 97 and note 12;
Prayon 1987, 29, 133, 185.

constructed some time after Tumulus W, whatever construction sequences

individual scholars choose. At issue up to the present is whether MM is the
second earliest after W, or actually the latest in the 8th century sequence
(see below), with III, IV, P and G in between.18 Some Gordion scholars con-
sider III, P and G to have been constructed close in time, that P follows III, or
that III is either earlier or later than IV, and so forth, but the precise relative
sequence is not necessarily significant here.19 In 1967 I noted that all the rel-
evant tumuli were constructed within a short space of time, perhaps within
one generation, but a ca. 50-year range should be considered, a range not
essentially inconsistent with the Gordion excavation teams assessment in
1981 and Sams later assessment.20 What is central here is that in all discus-
sions the Gordion team agrees (I too) that these tumuli were constructed
after Tumulus W, and sometime in the second half of the 8th century bc.
Moreover, and important for the present discussion, they have all agreed
that the chronology of several of the tumuli is not far from that of the DL.21
Based on parallels for the local and the imported artefacts from the
tumuli, members of the Gordion excavation team have argued both that
Tumulus MM was the last of the 8th century tumuli to have been con-
structed (contra Young), and that it was erected either immediately before
or after the destruction.22 Sams also considered Tumulus III and P to date
immediately prior to the DL, and contain materials that find their closest
parallels to DL artefacts.23 These archaeologically-generated observations
are critical for obtaining a conclusion about the DL chronology inasmuch
as they patently situate the chronology of the DL close to the construction
dates of MM (and III and P): to the late years of the 8th century bc. Archae-
ological consistency requires that if the DL occurred at the end of the 9th
century bc, the tumuli chronology must also be pulled back to that time;
the tumuli fibulae and imported material from MM are now documented
as artefacts of the first half of the 9th century bc. I suggest that the Gor-
dion team will not find this revised chronology easy to justify; they will have
to explain their previously argued 8th century bc tumuli-DL chronological
synchronism.
18 Respectively Young 1981, 241242; Sams 1994, 192.

19 DeVries, Sams, Mellink in Young 1981, 196, 198199, 270, 272; Kohler 1995, 38; Muscarella
1986, 195, 1998b, 183; Caner 1983, 7; Sams 1994, 4748, 215, 252, 256257.
20 Muscarella 1967, 4; Sams 1994, 4748, 215, 252, 256257.
21 Sams 1994, 187192, 195, and DeVries 1998, 8 for the latest claims.
22 Sams 1979, 4546; 1993, 551; 1994, 196; DeVries 1990, 388, 390the relation is clear; 1998,
8; Mellink 1981, 272.

23 Sams 1974, 169, 193.
I have dated Tumulus MM to ca. 720700bc, in part citing as argument 8th

century bc artefact parallels.24 Mellink dated MM to a post-DL time, DeVries
to a general early 7th century bc time, while Caner dates it to 730720 bc.25
The most recent dendrochronological adjustments (obviating earlier pub-
lished adjustments) fix the cutting of the tombs outer supporting logs at
740+4/-7 bcthus a range between 744 and 733 bc.26 But this cutting date
establishes only a post quem for the building of the tomb and deposition of
its artefacts. The building and furnishing of the tomb could have taken place
months, or one or more years after either 744 or 733bc, and it is appropri-
ate to recall again Kuniholms caution about the use of dendrochronology
to date structures at Gordion.27 The 744733bc range, higher than hitherto
supposed by all Gordion scholars (except Young), is not necessarily an issue
for the present discussion.28 It leaves unchanged the conclusion that all the
relevant tumuli are relatively close in construction dates (although perhaps
the new chronology might affect the view that Tumulus MM was the latest
constructed tumulus).
Tumulus S1 is the most difficult burial to date; it was plundered in the past
causing confusion about the recovered artefacts. At issue is its position with
regard to the other tumulipre- or post-DL; Kohler suggested that it is post
MM: neither had XII, 7A fibulae, and S1 had late fibula forms.29 Note that the
fibulae from the Mama Creek Gordion tumulus contained fibulae similar to
those from S1, and its date must be close.30 The 8th century bc Tumulus Q
has two fibulae that Caner considers to be XII, 7A, but which I consider to
be XII, 4 types.31
24 Muscarella 1967, 1, 3; 1982, 9; 1988b, 183; 1995, 99; 1998, 151153. For the imported objects
see Young 1981, 104110, 121123, pls. 5157, 6263; also Prayon 1987, 130132, 185, Taf, 41, 42.
None of these derive from Urartu, pace Mellink 1981, 268; 1991, 637.
25 Mellink 1981, 270; DeVries 1990, 390; Caner 1983, 9, 67.
26 James 2002, 18; Manning et al. 2001, 25332544.
27 Kuniholm 1988, 8.
28 This dendrochronological date would neatly fit Eusebius beginning year for the reign
of Midas at 738bcif the tomb were to have been constructed shortly after the cutting of
the logs. For years I have argued that Midas was not buried in Tumulus MM, that he built
the tumulus and tomb for a revered relative: see Muscarella 1967, 2; 1982, 9; 1988, 183; 1995,
99, note 13. This interpretation has been glossed over or ignored in publications and lectures
by the Gordion team who have often asserted that MM is the tomb of Midas. As of 2001 they
claim in memos and lectures that they now know for the first time ever that Midas was not
buried in MM.
29 Muscarella 1967, 6, 82; Caner 1983, 11; Sams 1994, 71, n. 64; Kohler 1995, 115121, and fig. 54.
30 Temizsoy 1992, pls. IX; 1993, pls. XIXVIIIesp. XII, 2, XII, 11/13A, XII, 14A, and fibulae
with a Gothic arcCaner 1993, pls. 1518.

31 Caner 1983, 17, nos. 188 B, C; Muscarella 1967, 15; 1986, 198; also Kohler 1995, 93, pl. 51 C,
D.
Fig. 2. XII, 7A fibula from Tumulus W (B 1260).
Fig. 3. XII, 7 fibula from Tumulus MM (B 1011).
Fig. 4. XII, 9 fibula from Tumulus MM (B 902).

Types XII, 7A, 7 and 9 fibulae appear in a number of tumuli at Gordion.

Type XII, 7A is found in Tumulus W (26 examples; Fig. 2), and in III, IV, S, and
G (one example in each); XII, 7A and XII, 7 occur together in Tumulus IV. All
scholars agree that XII, 7A and XII, 3, both with flat arcs, are the earliest Phry-
gian fibula types known; type XII, 3 occurs in W, III, and P, and together with
XII, 7A in Tumulus W and III. They also agree that XII, 7 is a modified, later
form of XII, 7A; it occurs in Tumulus MM (Fig. 3) and IV.32 Also accepted is
that XII, 9 is an embellishment, a later form of XII, 7; it occurs in tumuli III,
IV, MM (Fig. 4), and S1. Comparanda from other sites, including representa-
tions on reliefs at Khorsabad and Ivriz, support a late 8th century chronology
for both XII, 7 and 9 fibulae.33 XII, 13 occurs in a number of tumuli, W, III, IV,
P, MM, and S1, but those from W are clearly a distinct (early) form; XII, 14,
clearly a modification of XII, 13 and the latest in development among the
fibulae under review here, occurs in IV, MM, and S1. (There is discussion of
earlier and later forms of XII, 13 and 14 but these issues do not concern us
here.) Note that the Ankara tumulus of the late 8th century bc excavated
by Sevim Bulu contained the following fibulae: five XII, 3, thirty five XII,
7A, nine XII, 9, eighteen XII, 14 (information from Sevim Bulu); Caner only
mentions XII, 3, 7A, and 9, and presumably he was not allowed to publish
them; he dates the tumulus to ca. 750bc.34
The fibulae recovered from various structures in the DL consist of 24 Phry-

gian and 14 (possibly 15, see below) non-Phrygian, imported, examples. Of
the Phrygian fibulae, there are 19 XII, 7A examples (Fig. 5), one made from
electrum, two XII, 5one electrum, one silver (from the tumuli recov-
ered only in Tumulus IV), one very damaged XII, 9 (Megaron 4; Fig. 6; not
recorded in Caner), and one, XII, 14.35 One fibula fragment (B 1596) was
32 Muscarella 1967, 43, pls. IIIVI; see Appendix A for a chart of the tumuli loci of the
fibulae; 1998a, 426; 1988b, 179; Boehmer 1983, 7577, Abb. 1, 2.

33 Muscarella 1967, 1, 1719; 1982, 89; 1988a, 422, note 5, 426; Akurgal 1961, 61, fig. 38;
Boehmer 1983, 7679, Abb. 1, Taf. 21, 2.

34 Caner 1983, 15.
35 Muscarella 1967, pls. III, V. In personal correspondence DeVries informed me of a
discrepancy concerning the find spot within Megaron 4 of the XII, 9 fibula, B 1545: the
field notebook (written by the excavator) states it was recovered in the burned fill above
the Phrygian floor in cut 4Ca specific reference to a DL burnt debris locus; the inventory
record, however, states that it came from another area on the floor of Meg 4, cut 4S1. DeVries
believes that the clay overlay in the inventory recorded find spot (S1) seems undisturbed
but was probably [sic] churned up, while the clay overlying the notebook listed fibula (4C)
was badly disturbed by later scavenging. The implication seems to be to cast doubt on
the conclusion that the fibula derived from the DL, that it could have been dropped by
Fig. 5. XII, 7a fibula, destruction level (B 1361).
recovered from stone packing below the floor upon which Megaron 4 was
built; it preserves only the catch and section of the arc, round in form, sug-
gesting it may be a XII, 13 or 14 form; the arcs thinness suggest it is possibly
XII, 14.36 Megaron 4 was the latest structure to have been built before the
destruction, hence the fragment was deposited shortly before that event.
What is essential to recognise is that all the DL fibulae are the very same
types as those recovered in the 8th century bc tumuli W, III, IV, S, G and
MM, and include both early and late types.
In addition to the 14 imported examples from the DL, two others occur
in Tumulus G, and one in Tumulus Y, both 8th century tumuli. Eight of
these from the DL and those from Tumulus G and Y have swollen, leech
shaped arcs tapering to the ends and with flat catch plates (missing on the
Tumulus G example); slight rings occur at the ends of several examples.37
They are Aegean forms, Caners Type IVd, Sapouna-Sakellarakis Type IVc,
and, independently reinforcing the Gordion tumuli evidence, are dated to
a later intruderbut the absence of the studs, the damaged catch, and the missing pin
assuredly contradict this forced interpretation. Note that Sams (1994, 3) had no problems
with accepting a DL locus for the pottery he publishes from Megaron 4. The excavator
recorded disturbance resulting from the placement of the foundation of a MP building, and
did not doubt the DL origin of the fibula recorded in the field notebook. Fibula B 1454 derived
from the DL, along with EP pottery recorded here.
36 Caner 1983, no. 1179.
37 Caner 1983, nos. 84, 85, 88, 89, 92AD, 93; Muscarella 1967, pls. XVI, XVIII, figs. 8891
one example is not published; 1989, 336337, note 17; Kohler 1995, 87, pl. 21 C, D; 108109, pl. 57
A. Three imported fibulae found in the fill of Tumulus B are not counted here but should be
noted, see Muscarella 1967, 6, 82, Pl. XVIII, 92; Kohler 1995, 21f., Fig. 9BD, and dating the tomb
(p. 15) to the 7th century; Caner 1983, 181, no. 1191, dates it to the late 8thearly 7th century.
Caner 1983, no. 1184, a Near Eastern form, is not from the DL.
Fig. 6. XII, 9 fibula, destruction level (B 1454).
Fig. 7. Aegean fibulae, destruction level (B 1779).
the 8th7th centuries bc. (Mellink mistakenly claimed no Greek material

existed in the DL.)38
Three fibulae from Terrace Building 8 have a swollen arc framed by
prominent disc mouldings; one example (Fig. 7) has a hand catch, a typical
Near Eastern 8th century and later characteristic, here incised with an X.39
Note also that another example derived from the clay fill laid over the EP
and thus could have come to Gordion earlier than the MP period; if so, there
38 Caner 1983, 4143; Sapouna-Sakellarakis 1978, 7784; Mellink 1991, 631.

39 Caner 1983, nos. 1189, 1190 and 1192; Pedde 2000, 4, 7, Abb. 1, 2.
would then have been 15 imported fibulae in the DL.40 Caner catalogued
these fibulae as a central Anatolian variety of Near Eastern forms, Type V,
citing eleven close examples from the Phrygian period at Alishar; Pedde
categorised them in his broad Near Eastern C1.2 group.41 Fibulae of exactly
the same form also occur in central Anatolia at Kaman-Kalehykbut
unfortunately (as at Alishar) no information whatsoever exists in the site
publications concerning their loci and contexts that would help to situate
them within the local Phrygian sequence.42
Mouldings asymmetrically situated on the arc (B 1988c, dnot men-
tioned in Caner) characterise two of the imported fibulae in the DL; one
has a more pronounced asymmetrical moulding placement. I can find no
exact parallels, but Sapouna-Sakallerakis classifies them with her Type IV b
(all of which do not have asymmetrical arcs), allowing a range from Proto-
geometric (Vrokastro, her nos. 611 and 612which have a longer catch plate
than either of the Gordion examples) to Geometric.43 Giesen cites them as
Type III, Cypro-Archaic II; she sees them as possibly Cypriote but leaves
open a general Aegean distribution, and dates them to the late 8th cen-
tury bc.44
A fibula from Terrace Building 7 (B 1587; not in Caner 1983) is missing its
catch plate but has what appears to be an obliquely rilled or grooved arc. I
can find no exact parallels, but it seems similar to an example from the fill
over Tumulus Y, to fibulae from Tumulus S1, to the rilled XII, 13 arcs from the
Mama Creek tumulus and to a Phrygian fibula from 7th century Ephesus.45
It may belong to Peddes Group B2.3, with parallels from North Syriasee
especially his nos. 143, 146151, all dated to the 8th7th centuries bc.46
A review of the arc forms and mouldings of the relevant Phrygian fibulae
in the tumuli and DL reveals no clear indication of recognisable differ-
ences (other than in refinements) in the chronologies of their constructions.
Both Young and Caner (who classified all examples from the tumuli and DL
together within two subtypes) made the same judgment with regard to XII,
40 Caner 1983, no. 1193.

41 Caner 1983, 180188; Pedde 2000, 176, 180. In Muscarella 1967, 92, B2, I assumed these
were Aegean forms because of seemingly related examples from Rhodes.
42 For one publication of the Kamen-Kalehyk fibulae, see Mori and Omura 1993, fig. 12,
13, 14, Pl. 1.12.

43 Sapouna-Sakallerakis 1978, 7273; see her Taf. 9, no. 12 from Idalion.
44 Giesen 2001, 63, 109, 371.
45 Caner 1983, no. 1109; Muscarella 1967, pl. I, 1 and 2; Temizsoy 1992, pls. I, V, VIX.
46 Pedde 2000, 129133, 357.
7A fibulae (Figs. 2, 5).47 The DL XII, 9 fibula cannot be distinguished from

tumulus counterparts (Figs. 4, 6; Muscarella 1982, 8); it has a single rectan-
gular end moulding with studs, precisely as examples from Tumuli MM and
S1, but here the two end mouldings have rivets; a number of unexcavated
examples have one end moulding.48 The same conclusion obtains for the
XII, 14 example from the DL. Mellink thought this fibula was more devel-
oped than those in Tumulus MM (they occur also only in Tumulus IV and
S1)which would date it, in her (1981) understanding of the chronology of
MM, after 700 bc; Mellink also mentioned the XII, 7A fibula from the DL, but
not the XII, 9 example.49
To conclude, the Phrygian fibulae recovered in the DL and the tumuli
were manufactured within years or decades of each other. Judging from the
uncontested chronology of the tumuli this is a time within the second half
of the 8th century bcprecisely the chronology independently recognised
for all the foreign fibulae recovered in these same loci. Of parallel signif-
icance for chronological determinations is the fact that Phrygian fibulae
and locally made imitations have been excavated at sanctuaries all over the
Greek world, including Ischia in Italy: all dated to the 8th and 7th centuries
bc.50 Fibulae are important artefacts to be considered in archaeological
investigations of a sitehence the ample space accorded them in this study.
Fibulae do not occur anywhere in the inland areas of the ancient Near East
before the mid-8th century bc; they are a Leitfossil for this period and later,
as several scholars, recently Pedde, have demonstrated; this chronological-
geographical condition includes Phrygia and its capital Gordion.51 Fibulae
do not lie. If, as is the case, the same fibula forms known from the tumuli and
DL continued to be worn in the early phases of the MP, this is to be expected:
the rebuilders were Phrygians who of course continued to use Phrygian fibu-
lae. Put another way, their presence in the MP cannot be used per se to date
that period to the 8th, let alone the early 9th century bc.
47 Young 1981, 244; Caner 1983, 5960.

48 Caner 1983, Taf. 23, 2629, 33, 34; Taf. 31, 32; also Muscarella 1982, 8.
49 Mellink 1981, 269.
50 Muscarella 1967, 17, 1920, 2223, 26, 60, 63, pls. VI, VII, XI, Appendix C; 1989, 337339,
notes 2123; Boehmer 1983, 7577, Abb. 8.

51 For discussion and bibliography, see Muscarella 1984; 1988a, 425; Pedde 2000, 357359.
Fig. 8. Iron horsebit, destruction level (ILS 334).
Iron Horsebits
Three iron horsebits (Fig. 8) were found in the DL Terrace Building 2; the
cheek pieces are crescentic with a vertical central tang; rein holes exist on
the cheek piece and vertical tang. Young cited two examples, subsequently
noting that there should have been four of them because he thought they
were associated with the ivory horse trappings for four horses, see below;52
in fact, three were recorded (Keith DeVries and Gareth Darbyshire, personal
communications). I have been able to find formal parallels in two distinct
geographical areas. They occur at several sites on Crete, Knossos, Prinias,
and Arkades, all bronze.53 These examples consist of crescentic cheek pieces
with rein holes, but a much smaller vertical tang and the rein hole is below
it. The examples in the Knossos North Cemetery are dated from the pottery
to the Late Geometric or a little later. On a stone relief in a tomb at the
Transcaucasian site of Astchadzor was carved a related form, here with a
less cresentic formed cheekpiece, and a vertical tang, all apparently pierced
for reins; A. Ivantchik dates it to the 9th century bc, but his evidence is
not certain.54 According to I. Medvedskaya (personal communication) an
Assyrian 7th century bc-form sword was recovered in the tomb and she cites
52 Young 1962, 166167 pl. 48, fig. 26; Sams and Temizsoy 2000, no. 51.
53 Donder 1980, Taf. 8, 64, 65; Coldstream and Catling 1996, 213, Tomb 219, 210213, fig. 168,
nos. 92, 102, 103ahere too, as at Gordion, there is an odd piece.
54 Ivantchik 2001a, 144, 149, Abb. 67, 5.
two Russian scholars who respectively date the tomb to the 9th8th or the
7th centuries bc.
The bits found at Knossos occur together with a bronze bowl with lotus
handles, of the same form as one from Tumulus III and another of Egyptian
Blue from Tumulus P; other examples occur on Crete and in Cyprus, at Nim-
rud, Til Barsip, and Tell Halaf.55 G. Hoffman dates the Knossos (Tomb 219)
example to 745700bc, and another from Amnisos to 745710bc.56 Tumulus
P also contained Cypriote pottery, Black on Red II (IV), Cypro-Geometric III-
Archaic I vessels.57 Cypro-Geometric III is conventionally dated ca. 850
750bc, Cypro-Archaic I to 750600 bc, but some scholars place the chronol-
ogy higher, ca. 800750 bc (Joanna Smith, personal communication). In
either dating, high/low, we do not know how long it took for the vessels to
reach and then be deposited in a Gordion tomb.
It may be significant that the crescent form of the DL cheek pieces is
exactly matched by an apparent nose guard from a horse excavated in a
manifestly non-Phrygian (nomadic!) burial at Gordion, Tumulus KY, dated
by Kohler close to (after?) the DL.58 Here the very same crescent shape is
formed as two units joined together as one piece. What this clear parallel
signifies, in both cases involving horse equipment, I leave open: except to
note that the certain late (late 8thearly 7th century) chronology of Tumulus
KY is to be kept in mind when dating the iron bits.
The iron horsebits from the DL most probably were manufactured some-
time in the 8th century bc. Sams believed (as did Young, above) that they and
the ivory horse trappings (see below) also found in TB 2, were used together
and were immediately recognised as North Syrian products (but he gave
no parallels for the horsebits) and came to Gordion as a unit through a
diplomatic exchange (along with cauldrons); earlier he identified the North
Syrian centre as Carchemish.59 Indeed, the ivories very probably came from
North Syria, but I suggest that the iron horsebits probably derived from
another area. One may viably speculate that the horsebits arrived at Gor-
dion along with their horse(s), making an overland trip within Anatolia in
the 8th century bcbut from which direction?
55 Muscarella 1967, 68; Young 1981, 31; Hoffman 2000, 3637.

56 Hoffman 2000, 36, 134.
57 Young 1981, pl. 14, D, E, pl. 16, D, pl. 17, H.
58 Kohler 1995, 80, KY 46 C.
59 Sams 1993, 552553; 1979, 46.
Ivory Frontlets and Blinkers
Four ivory frontlets and five (Young thought there should have been eight,
for four horses) cheekpieces or blinkers (Figs. 910) as noted above were also
found in Terrace Building 2 near the iron horsebits.60 The frontlets, some very
fragmented, are all decorated with a winged solar symbol set above a frontal-
nude female who holds two winged sphinxes by their rear legs. The blinkers
all show a chimera, a winged lion with an en face human head projecting
from the shoulder, and a tail ending in a birds head, typical North Syrian
motifs of the 8th century bc.61 Add to this that the ivory lions head muzzle,
ear form and lolling tongue are also neatly paralleled on a Phrygian painted
lion that Sams assigned to the EP citadel; and a sphinx depicted on a vessel
from Tumulus P has a bird-ending tail.62
With the high chronology these DL ivories (and also their pottery motif
counterparts?) are now assumed to be imports from 9th century North
Syriaan assumption contradicting M. Voigt and R. Henrickson who, invok-
ing North Syrian parallels, dated the ivories to the 8th century bc.63 How-
ever, now brought forth as 9th century parallels for the ivory frontlet is a
bronze frontlet excavated on Samos, and, for the blinkers, two bronze exam-
ples excavated in Eretria, Greece (Voigt and DeVries, personal communica-
tions); the Samos frontlet and one of the Eretria blinkers bear inscriptions of
Haza"el, presumably the 9th century King of Damascus.64 But if one but looks
at the Gordion ivories and their alleged chronological parallels one instantly
recognises that the Eretria blinkers show a male, not a female, holding lions
by their hind legs; there is no chimera, and they have different end tang
shapes. One also sees that the Samos frontlet exhibits a different iconog-
raphy than the Gordion examples, with three nude females holding their
breasts, above a fourth. Further, all faces, mouths, eyes, hands, and hair, and
bracelets, anklets and necklaces, differ notably in style and body decoration
from those in the Gordion examples.
60 Young 1962, 166167, pls. 46, 47, figs. 24, 25; Sheftel 1974, 408425, pls. 6062; Prayon 1987,
184185, Abb. 30, Taf. 44 c, d.

61 Sheftel 1974, 422424; Sams 1974, 187188, 191; 1979, 46; 1993, 552; Donder 1980, 28,
note 29, 103; DeVries 1981, 221; Prayon 1987, 183186.

62 Sams 1994, 88, no. 1077, pl. 131; Young 1981, 36, fig. 19.
63 Voigt and Henrickson 2000, 50. Sheftel (1974, 419, 421425) makes strong arguments that
the Gordion pieces derived from North Syria.

64 Muscarella 1970, 116, fig. 10; Donder 1980, nos. 194198, 201 and 202; Eph"al and Naveh
1989, pls. 24, 25 (see also Addendum, below).

Fig. 9. Ivory horse frontlet, destruction level (BI 432).

Fig. 10. Ivory horse blinker, destruction level (BI 440).
The winged sun forms on the Samos bronze and the Gordion ivory front-
lets are entirely different, except that their discs have an interior crossed
motif, a rosette in the former, crossed lines in the latter. This interior design
does not appear on any of the Nimrud ivories but it does occur on the reliefs
of two 8th century sites in North Syria, Sakcegz and Zincirli. And it is on
Nimrud ivories of manifest 8th century date, not the Samos frontlet, that
exhibit the very same pendent palm below the sun disc.65
As for the date of the loci in which the alleged parallels occur, it is
noteworthy that the inscribed Samos frontlet was recovered in a 6th century
deposit, the Eretrian blinkers in a late 8th century stratum. What we have
here are objects apparently dated to the 9th century by inscriptions that
were recovered in levels dated hundreds of years laterall of which is not
necessarily relevant to the Gordion ivory pieces anyway, for the alleged
parallels have nothing stylistically in common with them. At present only
one site that may be brought into a discussion for formal and iconographical
parallels on the Gordion frontlets, and that is Nimrud, a site destroyed in the
late 7th century bc. Here as at Gordion are ivory frontlets depicting a frontal
nude female holding lions by their tails, and in at least one example framed
by a guilloche.66
65 Orthmann 1971, pl. 49, A/1, 50, A, 10, pl. 66, K/2; Sheftel 1974, 196197.
66 Sheftel 1974, 417; Orchard 1967, pls. XXVIIIXXXI; Prayon 1987, 185186.
Fig. 11. Glazed juglet, destruction level (G 224).
Imported Pottery
A glazed juglet (Fig. 11) of Syro-Palestinian form was found inside a vessel
in Terrace Building 4.67 The juglet is a classic Iron II shape and can be
dated to the late 8th century bc, continuing into the 7th6th centuries
(Larry Herr and Bruce Routlege, personal communications)which neatly
informs us that it could not even have been a gleam in a potters eye in the
9th century bc.
There is also a small corpus of Proto-Corinthian sherds, not from the
DL, but some from the citadel. Although not from a stratigraphically infor-
mative context, DeVries assigned them to the EP period, assuming they
were picked up in pit digging; Sams, noting that they were recovered scat-
tered and dislocated, also concluded that they probably derived from the EP
level.68 Voigt and Henrickson disagreed, stating that the sherds (incorrectly
described by them as semi-complete) were recovered in the MP period.
They dated the sherds to late 8th or early 7th century; Sams dated them to
about 725720; Mellink to 720690 bc.69 The issue here is that some Gordion
67 Sams 1993, 552, pl. 94, 2; 1994, 64, 271, pl. 86, no. 786.
68 DeVries 1990, 390, fig. 25; Sams 1979, 47, fig. 3; 1994, 135; also Mellink 1991, 631, note 27.
69 Voigt and Henrickson 2000, 52; Mellink 1981, 269.
scholars interpret them as originally imported during the EP period (see the
stone reliefs, below), others that they were imported in the MP period.
The original loci of the sherds will never be determined, nor perhaps
may the lack of a more precise chronology for the terminal dates assigned
the sherds be resolved. But if the terminal date of ca. 690bc or even a
decade earlier holds up (i.e. 700 +/- bc), one is aware that each interpretation
regarding their original locus equally suggests/indicates either a date for the
time for the D1 or the time when the MP was coming to be or was already in
existenceand that is the late 8th or early 7th century bc.
Bulls Head Attachments
Three bronze bulls head attachments were recovered from the DL, two in
Terrace Building 1, one in Megaron 4; the two from TB 1 are a pair certainly
from a cauldron, which leaves one missing from Megaron 4 (see Note 35).70
They relate stylistically and structurally to the bulls head attachments on
cauldrons from Tumulus W and MM of the 8th century bc,71 a date recog-
nised by DeVries and Muscarella for these and other bulls head attachments
in the Near East.
Orthostate Reliefs
Sams (1989) presented a good summary of several of the ten fragmentary,

locally carved, reliefs that derived from rubble associated with the building
of MP structures. He persuasively interpreted them to have been salvaged
from the EP level, and also showed that their architectural background is
to be found in North Syria. He presented 10th9th century parallels from
that area, in particular from Carchemish and Zincirli. Speaking to the date
when they might have been carved at Gordion, he cogently noted that there
is no ready answer. But because the North Syrian examples were visible and
could have been available as examples through the eighth century, even a
time shortly before the destruction, and because there is historical docu-
mentation of contact between Phrygia and Carchemish at this period, the
8th century could have been the time of the Gordion carving. Nevertheless,
he raised the possibility, which he chose to accept, that they could have been
70 DeVries 1981, 221222, pl. 95 B, D; Muscarella 1989, 340341.

71 Young 1981, pls. 59, AC, 88, A.
copied at a time that approaches the date of the parallel material and thus
be as early as the ninth century, no later than 800 bc.72
I argue with ease that awareness about the period of time available for
visual knowledge of and accessibility to the North Syrian reliefs is the objec-
tive and correct archaeological position, and this precludes a 9th century
conclusion.73 Furthermore, as Sams admits, it was in the late 8th century,
and not earlier, that we have information about mutual contacts. Samss
9th century date is not a substantive conclusion; it is but a guess (compare
his comments above about 8th century bc iron horsebits and other mate-
rial coming from North Syria in the late 8th century). Voigt and Henrickson
handled Samss 9th century date as an historical fact, which it patently is
not; their acceptance was unexamined.74 And this acceptance cannot be
invoked as evidence either to support a 9th century bc date for the local
carving of the reliefs, or as evidence for the date of the DL. One should recall
here Samss discussion of North Syrian influences on figures represented on
Gordion pottery, dated by him to the 8th century bc. He suggested that the
motifs could have been adapted from North Syrian reliefs that were visible
at least until the end of the eighth century or even the early 7th century bc.75
Griffin Vessel Handle (Fig. 12)
From Megaron 3 was recovered a single vessel handle attachment in the

form of a winged, open-mouthed griffin, its feet pulled up against its body, as
if seen from below, and with large, horn-like, ears, and a knob on the top of
its head.76 Both R. Young and F. Prayon called attention to its resemblance to
two late 9th century bc winged bird-headed vessel handles on a bowl from
Hasanlu in north-western Iran, especially relevant being the pulled up feet
placement.77 The Gordion griffins feet position and its relationship to the
Hasanlu bird examples warrant attention, for the question has been raised
whether the former is also 9th century in date [see Addendum below]. I
suggest this is not defensible. The Hasanlu attachments are birds, and no
griffin heads are known from that site, while bird-headed attachment
72 Sams 1989, 452453; but not in 1994, 194.

73 See also Mellink 1991, 641.
74 Voigt and Henrickson 2000, 50; but Voigt 1994, 272 misquotes him.
75 Sams 1974, 183184, 187.
76 Young 1962, 163, pl. 43, fig. 15; Porada 1962, 5354, pl. 9, right.
77 Prayon 1987, 127128, Taf. 20 d; Muscarella 1970, 115, fig. 9; fig. 8 is the Gordion example.
Fig. 12. Griffin vessel attachment, destruction level (B 1332).
but with no legs shownoccur in Tumulus W.78 As I noted earlier, the

Gordion griffin in fact has close parallels with an open-mouthed griffin
vessel attachment from Olympia in Greece, of the 8th century bc or later.79
Further, the knob on the Gordion griffin head together with its open mouth
are typical features of North Syrian and Greek griffin heads, which indicate
a non-Iranian background.
However, a gold raptor, an ornament not an attachment, with its feet in
the same pulled up position, occurs at Susa, possibly from the Old Elamite
period.80 But the same form exists elsewhere, for example at Ephesus, of the
7th century bc as a gold brooch, and at Nimrud (in a late 7th century bc
destruction context!) in the form of double-ram vessel attachments, their
feet held up before them.81 The specific origin and exact date of the Gordion
attachment is not to be dated solely by the Hasanlu attachments. It is also of
value to recall that two bronze griffin head protomes of typical Greek form
were recovered in the Phrygian tumulus excavated near Bayndr of late 8th
century date (Antalya Museum Catalogue 1988, nos. 29, 30).
78 Young 1981, 201202, fig. 118, pl. 88, B, C.

79 Muscarella 1970, 114, fig. 7; see also Prayon 1987, 128129.
80 Porada 1962, 5354, pl. 9, left.
81 Akurgal 1961, 210, fig. 174; Layard 1853, 185.
Phrygian Inscriptions
I am not a linguist and approach this issue solely on archaeological grounds.

One Phrygian inscription (no. 29) was recovered below the floor of Megaron
10 and is thus manifestly earlier in date than the DL itself.82 How earlier is
of course a problem, but all attempts to arrive at a solution are subjective
assumptions not based on firm evidence. Young claimed it should be dated
to 750bc or perhaps earlier (i.e. ca. 50 years before the destruction), and
thus is the earliest Phrygian inscription known. But his dating is conjecture,
based solely on how long he estimated for the floor to be laid in place over
the inscription. No archaeologist knows how many years the inscription lay
in the ground before the destruction, for there is no chronologically deter-
mined stratigraphic evidence. Nevertheless, it is noteworthy that, contra
Young, DeVries considers the below-floor stratum containing the inscrip-
tion to be dated to perhaps the second half of the eighth century, i.e. closer
to the time of the DL.83
C. Brixhe accepted Youngs chronology without discussion, also claiming
it to be the earliest Phrygian inscription known. He further asserted, but
without presentation of evidence or explanation, that from the no. 29 locus,
it is reasonable to date the beginning of the Phrygian alphabet to the late 9th
or early 8th century bc.84 M. Lejeune also accepted Youngs chronology, and,
as did Brixhe, concluded that the Phrygian alphabet developed in the first
half of the 8th century bc.85 Both conclusions are suppositions, not based
on evidence at Gordion. Philologists, like archaeologists, create their own
chronologies.
Another inscription, on a stone (no. 39), originally must have been in use
before the DL, as it was recovered in the rubble foundation of a MP build-
ing.86 In addition to these (rare) DL inscriptions, five also were deposited
in Tumulus MM.87 I have previously argued and affirm again that there
is no archaeological evidence from Gordion that demands the conclusion
that the Phrygian alphabet can be dated to 750 bc or a decade or more
82 Young 1969, 256258, no. 29.

83 DeVries 1990, 390.
84 Brixhe 1981, 275.
85 Lejeuene 1970, 64; but see his criticism of a dating by Young of another excavated
inscription at Gordion, p. 53.

86 Young 1969, 271, no. 39.
87 Young 1969, 259262; note that the inscriptions illustrated in Sams and Temizsoy 2000,
nos. 6063, do not derive from the pre-DL period as stated there.
earlierinvoking here the low chronology of course.88 Further research

must take place before it may be determined whether the DL inscriptions
or the Tumulus MM inscriptionspost-744/733bcare the earliest Phry-
gian inscriptions recorded to date: the precise chronological relationship of
the DL and the earlier Megaron 10 inscriptions to each other and both to the
deposition date of Tumulus MM remain unanswered questions (and what
relationship can be postulated about the Gordion inscriptions chrono-
logical relationship to those of the Midas Monument and Tyana inscrip-
tions?).
If the high DL chronology is accepted as historical reality the 8th cen-
tury bc date of the Tumulus MM inscriptions is not affected (but see above);
however, all the hitherto postulated dates jump for the DL, thus signifying
that the Phrygian alphabet was in existence in the 9th century bc.
This is a thought-provoking revision of all that is known about the floruit
of the Phrygian alphabet. But it also radically affects the chronology of the
incipience of the Greek alphabet. Young considered the Phrygian and Greek
alphabets to have derived independently of each other from a common
ancestral alphabetat the same time admitting, that the vowels were not
independently derived, and even (reluctantly?) accepting a Greek chrono-
logical priority.89 Lejeune, contra Young, and together with a long list of
scholarsincluding DeVries and Samsbelieve that the Phrygian alpha-
bet did indeed develop from the Greek.90 An important issue to reflect upon
is that the Phrygians chose to borrow a Greek script, not the cuneiform used
in Assyria, or the cuneiform used in Urartu (borrowed from Assyria in the
late 9th century bc), or the Luwian hieroglyphic used in North Syria. The
choice surely reflected close political and cultural relations with the Greek
sphere to the West in the 8th (sic.) century bc.91
The earliest Greek inscriptions are dated to post-750 bc. In the high chro-
nology, the Phrygians would have developed their alphabet more than 50 to
80 years before the Greeks did; and if Youngs 50-year estimation about the
internal chronology for Megaron 10 is accepted, we are talking of a century
or more. I suggest that such a chronology would strike Greek philologist
and I suggest some Phrygian philologistsas more than extraordinary. And
all the more so for Brixhes and Lejeunes posited estimations for the early
incipience of the Phrygian alphabet, which would now inevitably have to
88 Muscarella 1967, 8.
89 Young 1969, 255, 265, 296.
90 Lejeune 1970, 64; DeVries 1990, 64; Sams 1994, 176.
91 Muscarella 1989, passim, 337; see also above and note 50.
jump to the late 10thearly 9th centuries bc. I argue that the archaeology of
Gordion does not allow acceptance of these important conclusions.
Cimmerians
A comment on the Cimmerian involvement in the destruction of Gordion

is relevant to the issue of chronology. If the DL was destroyed in the late
9th century bc the Cimmerians could have had nothing to do with the
destruction, end of the discussion; but if the generally accepted chronolog-
ical range for the DL date (ca. 700 to 676 bc) holds, as I trust I have made
clear that it should, then the Cimmerians must be brought into the dis-
cussion. Cogent discussions of the issues of the destruction chronology at
Gordion and the issue of Cimmerians in Phrygia are brought forth both by
Ivantchik, who dates the destruction to the Cimmerians in 670 bc, and E.-
M. Bossert, who dates the same Cimmerian event to 674 bc.92 S. Tokhtasev
said it quite accurately; we cannot date either Midas death or the Cimme-
rian invasion more precisely than 700675 bc.93 Not to be overlooked is that
Ivantchik and Bossert disagree only in a matter of years with the traditional
recorded dates of Eusebius and Julius Africanus. Both scholars also judge
the destruction to have occurred during the reign of the Mita-Midas men-
tioned by Essarhadon in texts dating to the 670salthough no destruction
is mentioned here. This Midas, called Lord of the City, not kingbut which
term could have been merely dismissive, and not, as Bossert believes, indi-
cating that Phrygia was an Assyrian vassalwas probably a successor of the
MitaMidas mentioned by Sargon II in 717 and 709bc, perhaps his grand-
son, who would have reigned in the MP citadel.
D. Hawkins suggests that the Cimmerians were in Anatolia in 709bc,
and that Sargon II was killed in a battle against them in Tabal; DeVries
rightly rejects this assumption, which cannot at present be demonstrated as
historical reality.94 I also reject, because it is not demonstrated, being based
solely on a restored text (SAA IV: 1), the view that in the 670s the Cimmerians
and Phrygians were allies. Many scholars choose to accept the alliance, but
Tokhtasev again put it accurately: the alliance is conjectural.95
92 Ivantchik 1993, 6774; Bossert 1993, 289291.

93 Tokhtasev 19911992, 565.
94 Hawkins 1982, 420, 422; DeVries 1998, 67.
95 Viz. Ivantchik 1993, 68, 74; Bossert 1993, 291; Kohler 1995, 76; DeVries 1998, 7; Tokhtasev
19911992. 565.
Voigt and Henrickson tell us that the ancient historians who refer to
Cimmerians and Phrygia lived hundreds of years after the putative events
and therefore cannot inform us about who destroyed Phrygia (Gordion).96
Only modern scholars, who are more competent and knowledgeable, can
do this correctly. But it is a matter of fact that the judgments of these
ancient historians, based on ancient texts available to them, were not far
from historical chronologies and data known to modern historians from
ancient Assyrian texts. Ivantchik persuasively notes that we cannot claim
to have in our possession all the original Assyrian documents, especially
letters to and from kings, that could have informed us whether or not
the Cimmerians were or were not in Anatolia between 714when they
are first mentioned, in north-western Iranand the 670s; our records are
incomplete.97 Decidedly pertinent in this context is a stone inscription from
Samos dated to 283282bc that is fittingly cited by Ivantchik, for it sheds
valuable light on the issue of ancient knowledge of the Cimmerians.98 The
inscription records a boundary dispute between Samos and Priene, a city in
western Anatolia, and refers specifically to earlier, preserved and archived
documents reporting the Cimmerian invasion in the mid 7th century bc,
i.e. 400 years earlier. This inscription neatly documents for modern scholars
the fact of ancient record keeping by people who lived hundreds of years
after the [not putative] events.
Voigt and Henrickson cite DeVries as one who has convincingly elimi-
nated the Cimmerians as the culprit.99 This strong, but actually unsubstan-
tive claim is based solely on DeVries dating of the DL to 710bc, and naming
Sargon II as the culprit. But, albeit accepted by three Gordion scholars (but
only for a short time, until Heidelberg changed their minds) the 710 date is
a modern fabrication: no such event is recorded in any ancient, including
Assyrian, texts. We are left after all with the Cimmerians as the only histori-
cally documented invaders of Phrygia.
Voigt and Henrickson not only reject the Cimmerians as the culprits, they
also assert there was little evidence for a battle, and deny that it was a battle
that caused the DL in the first place (T. Cuyler Young agrees with this assess-
ment, personal communications). To Voigt, only the graphic evidence from
Hasanlu IV in north-western Iran, with its over 200 dead in the burnt and
96 Voigt and Henrickson 2000, 51.

97 Ivantchik 2001b, 323, 331332.
98 Ivantchik 1993, 113114; 2001b, 323.
99 Voigt and Henrickson 2000, 52 (ignoring DeVriess contrary view expressed in DeVries
1990, 388, 390); DeVries 1998, 7.

collapsed debris, qualifies as an archaeologically correct interpretation for a

destruction caused by battle. This is a skewed conclusion, for if such were the
case, then hundreds of destruction levelsburnt and abandoned sites
all over the Near East where no, or few, bodies were recovered are ipso facto
not battle-caused destruction levels. Hasanlu IV is of course a unique phe-
nomenon, and Voigt makes no mention of other similar situations. DeVries
presented a contrary, and to me accurate, conclusion: the fire seems
not to have been accidental, but a deliberate enemy destruction, as is
indicated by the pattern of burning (for the extent of the destruction see
also an earlier comment of Voigt herself).100 The DL indisputably reflects a
major destruction, and surely one caused by an enemy attack. As with many
enemy-caused destructions, the site was abandoned, the population prob-
ably surrendering to the attackershence the absence of (human) bodies,
and the presence of the debris littered ruins and the many possessions not
salvaged, all covered over when the MP citadel was built. Identifying the
enemy of course remains a matter subject to debate, but the ancient sources
that address the destruction of Phrygia must be considered, not arbitrar-
ily ignored or rejected. Nomadic destructions do not leave many material
remains; the attackers at Gordion left no calling card; they came, conquered,
and left.101
To conclude: we have before us a chronology based on a C 14 analysis that

indicates the artefacts in the DL, but not those in the burial tumuli that share
a close chronological connection, are to be dated to the late 9th century bc.
And parallel to this situation we have a review of the chronology of the
artefacts in the DL and the tumuli that (re)affirms a shared date for both
units, the second half of the 8th century bc (whether terminating in 696 or
later). We have before us a conundrum; something is wrong here. I suggest
that the problem lies not in the artefacts, which indicate that Gordion was
destroyed sometime about 700bc or slightly later.102
100 DeVries 1990, 388; 1997, 1; Voigt 1994, 272273.

101 See Ivantchik 2001b, 334; Mellink 1991, 634.
102 P. Magee (The Iranian Iron Age and the Chronology of Settlement in Southeastern Ara-
bia, IrAn XXXII, 1997, 91108) rightly rejects the high C 14 determined chronology accepted
by the excavators of Rumeilah in Arabia (14th9th centuries bc) because of much later-dated
Iranian influenced pottery recovered there (fig. 2, B, C, E, G). But Magee in turn dates this
potteryand thus the sitestill too high by probably a hundred years; it probably existed
not earlier than the 8th century bc (painted vessels from Sialk are not manifestly 9th cen-
tury bc). Moreover, Magee himself at his site Muweilah in the Emirate of Sharjah enthusi-
astically accepts here a C 14-based initial date of 10th/9th century bc (see for one report:
Addendum
In June 2003, after the present paper had been accepted for publication,
I read on the internet an article by K. DeVries, P. Kuniholm, K. Sams and
M. Voigt, New Dates for Iron Age Gordion, published in that months issue
of Antiquity. The article is short but mentions the C 14 dates extending from
824807bc , and it singles out as independently support[ing] the late
9th century date the ivory horse trappings and the bronze griffin handle
(called a protome), citing Haza"el horse trappings and the Hasanlu bird
handle (protome) as the relevant chronological parallels. Fibulae and other
artefacts are not mentioned. The authors assert that their conclusions indi-
cate a severe gap between Phrygian material culture and the more modest
culture of Geometric Greece . The present paper, I suggest, demonstrates
that their conclusions both on the date of the Gordion artefacts and a now
revealed cultural unbalance of an alleged 9th-century Phrygia with that of
Greece cannot be accepted.
Acknowledgments
It was Rodney S. Young who first introduced me to the extraordinary site

of Gordion and allowed me to witness the excavation of Tumulus MM. For
that gift (and more) I have never forgotten him. I thank Elizabeth Simpson
for her rigorous reading and critique of this paper; T. Cuyler Young and
Maya Vassileva for close readings and discussions about its contents; Maya
Vassileva for discussions about fibula and other issues and chronology; and
Mary Voigt and Keith DeVries for vigorously sharing with me their opinions.
Keith DeVries and The University of Pennsylvania, and Maya Vassileva,
generously gave me with a number of photographs published here.
Excavations at Muweilah 19972000, Proc. of the Seminar for Arabian Studies 31, 2001, 115
130). And he claims as chronological support bridge-spouted vessels recovered there, asserted
by him to be Iron II forms9th century bc. In fact, the Muweilah vessels have vertical, not
horizontal spouts (ibid., fig. 12), a characteristic not of Iron II, but of later vessels, dating after
800 bc (when Hasanlu IV was destroyed). Further, to seek a model for the columned hall exca-
vated at Muweilah one need not with the excavator look to far away 9th century bc Hasanlu
for a direct source: exact and geographically closer models existed in Iran in the 7th and 6th
centuries bc at Nush-I Jan, Godin II, Pasargadae, and Persepolis (Muscarella 1988a, 208209,
note 3). Magee should not automatically have accepted the C 14 date assigned to Muweilah
for the very same reasons he argued against accepting the C 14 date assigned to Rumeilah!
Bibliography
Abbreviations
ACSS Ancient Civilizations from Scythia to Siberia
AJA American Journal of Archaeology
AnatSt Anatolian Studies
BASOR Bulletin of the American Schools of Oriental Research
IEJ Israel Exploration Journal
IrAn Iranica Antiqua
IstM Istanbuller Mitteilungen
JFA Journal of Field Archaeology
QRArch Quarterly Review of Archaeology
Akurgal, E. 1961: Die Kunst Anatoliens (Berlin).

Boehmer, R.M. 1983: Eine tabalische XII 9 a Fibel aus Assur. In Boehmer, R.M. et al.
(eds.), Beitrge zur Altertumskunde Kleinasiens (Berlin), 7582.
Bossert, E.-M. 1993: Zum Datum der Zerstrung des phrygischen Gordion. IsM 43,
287291.
Brixhe, C. 1983: Inscriptions. In Young 1981, 273277.
Caner, E. 1983: Fibeln in Anatolien I (Munich).
Coldstream J.N. and Catling, H.W. 1996: Knossos North Cemetery: Early Greek Tombs
(London).
DeVries, K. 1981: Comments in Young 1981.
. 1990: The Gordion Excavation Seasons of 19691973 and Subsequent Re-
search. AJA 94, 3, 371406.
. 1998: The Assyrian Destruction of Gordion. Paper delivered in Chicago at
the Annual Meeting of the Archaeological Institute of America.
Donder, H. 1980: Zamzeug in Griechenland und Cypern (Munich).
Eph"al, I. and Naveh, J. 1989: Hazaels Booty Inscriptions. IEJ 39, 34, 192200.
Giesen, K. 2001: Zyprische Fibeln: Typologie und Chronologie (Jonsered).
Hawkins, D. 1982: The Neo-Hittite States in Syria and Anatolia. In Cambridge An-
cient History, III.1 (Cambridge), 372441.
Hoffman, G.L. 2000: Imports and Immigrants (Ann Arbor).
Ivantchik, A. 1993: Les Cimmirians au Proche-Orient (Gttingen).
. 2001a: Kimmerien und Skythen (Moscow).
. 2001b: The Current State of the Cimmerian Problem. ACSS 7, 34, 307339.
James, P. 2002: The Dendrochronological Debate. Minerva XIII, 4, 18.
Kohler, E. 1981: Comments in Young 1981.
. 1995: The Lesser Phrygian Tumuli, Part I, The Inhumations (Philadelphia).
Kuniholm, P.I. 1988: Dendrochronology and Radiocarbon Dates for Gordion and
other Phrygian Sites. SOURCE VII, 34, 58.
Layard, A.H. 1853: Discoveries in the Ruins of Nineveh and Babylon (New York).
Lejeune, M. 1970: Les Inscriptions de Gordion et lalphabet phrygien. Kadmos 9,
5174.
Manning S.W., Kromer, B. et al. 2001: Anatolian Tree Rings and a New Chronology
for the East Mediterranean Bronze-Iron Ages. Science 294, 25322535.
Mellink, M. 1981: Comments in Young 1981.

. 1991: The Native Kingdoms of Anatolia. In Cambridge Ancient History III.2
(Cambridge), 619665.
Mori, M. and Omura, S. 1993: A Preliminary Report on the Third Excavation at
Kaman-Kalehyk in Turkey (1988). In Makisa, T. (ed.), Essays on Anatolian
Archaeology (Wiesbaden), 4374.
Muscarella, O.W. 1967: Phrygian Fibulae from Gordion (London).
. 1970: Near Eastern Bronzes in the West. In Doeringer, S. et al. (eds.), Art and
Technology (Cambridge), 109128.
. 1982: King Midas Tumuius at Gordion. QRArch 34, 710.
. 1984: Fibulae and Chronology, Marlik and Assur. JFA 11, 413419.
. 1986: Review of Caner 1983, Biblioteca Orientalis 12, 194202.
. 1988a: Bronze and Iron, Ancient Near Eastern Artifacts in the Metropolitan
Museum of Art (New York).
. 1988b: The Background of the Phrygian Bronze Industry. In Curtis, J. (ed.),
Bronze-working Centres of Western Asia. 1000539BC (London), 177192.
. 1989: King Midas of Phrygia and the Greeks. In Emre, K. et al. (eds.), Anatolia
and the Ancient Near East: Studies in Honor of Tahsin zg (Ankara), 333
344.
. 1992: Greek and Oriental Cauldron Attachments: A Review. In Kopcke, G.
and Tokumaru, I. (eds.), Greece Between East and West 10th8th Centuries bc
(Mainz), 1645.
. 1995: The Iron Age Background to the Formation of the Phrygian State.
BASOR 299/300, 91101.
. 1998: Relations Between Phrygia and Assyria in the Early Eighth Century BC.
In XXXIVme Rencontre Assyriologie International, Ankara 1987 (Ankara), 149157.
Orchard, J.J. 1967: Equestrian Bridle-Harness Ornaments (London).
Orthmann, W. 1971: Untersuchungen zur spathethischen Kunst (Bonn).
Pedde, F. 2000: Vorderasiatische Fibeln von der Levante bis Iran (Saarbrcken).
Porada, E. 1962: The Art of Ancient Iran (New York).
Prayon, F. 1987: Phrygische Plastik (Tbingen).
Sams, K. 1974: Phrygian Painted Animals: Anatolian Orientalizing Art. AnatSt XXIV,
169196.
. 1979: Patterns of Trade in first Millennium Gordion. Archaeological News 8,
4553.
. 1981: Comments in Young 1981.
. 1989: Sculpted Orthostates at Gordion. In Emre, K. et al. (eds.), Anatolia and
the Ancient Near East: Studies in Honor of Tahsin zg (Ankara), 447454.
. 1993: Gordion and the Near East in the Early Phrygian Period. In Mellink, M.
et al. (eds.), Aspects of Art and Iconography: Anatolia and its Neighbors. Studies in
Honor of Nimet zg (Ankara), 549555.
. 1994: The Early Phrygian Pottery (Philadelphia).
Sams, K. and Temizsoy, I. 2000: Gordion Museum (Ankara).
Sapouibna-Sakellarakis, E. 1978: Der Fibeln der griechischen Iseln (Munich).
Sheftel, P. 1974: The Ivory, Bone and Shell Objects from Gordion from the Campaigns
of 1950 through 1973, PhD Dissertation (University of Pennsylvania).
Strobel, K. 2002: Menschenopfer und Kannibalismus. Antike Welt 33, 487491.
Temizsoy, I. 1992: The Excavation of the tumulus at Mamaderesi. Anadolu Medeni-

yetleri Muzesi, 127.
. 1993: The Mamaderesi tumulus Excavations, Small Finds II. Anadolu Medeni-
yetleri Muzesi, 553.
Tokhtasev, S.R. 19911992: Cimmerians. In Encyclopaedia Iranica 5 (New York), 563
567.
Voigt, M. 1994: Excavations at Gordion 19881989: The Yassihyk Stratigraphic
Sequence. In inaroglu, A. and French, D. (eds.), Anatolian Iron Ages 3 (Ankara),
265293.
Voigt, M. and Henrickson, R. 2000: Formation of the Phrygian state: the Early Iron
Age at Gordion. AnatSt 50, 3754.
Voigt, M. and Young, T.C. 1999: From Phrygian Capital to Achaemenid Entrepot:
Middle and Late Phrygian Gordion. IrAn XXXIV, 191241.
Young, R.S. 1962: The 1961 Campaign at Gordion. AJA 66, 153168.
. 1969: Old Phrygian Inscriptions from Gordion: Toward a History of the
Phrygian Alphabet. Hesperia 38, 252296.
. 1981: Three Great Early Tumuli (Philadelphia).
chapter nineteen
AGAIN GORDIONS EARLY PHRYGIAN DESTRUCTION DATE:

CA. 700 +/- BC*
Abtsract: It is a pleasure for me to write a paper on an important issue in Anatolian

archaeology for my dear friend and colleague Aykut Cinaroglu, whom I first met in
1984 when he had a Fellowship at the Metropolitan Museum of Art (later in Ankara
I met often his wonderful wife and daughters and their sweet dog, Tarin).
Introduction
This paper continues the discussion presented in Muscarella 2003 (see also
Muscarella 2005/2006: 395, and note 4) concerning the date of the destruc-
tion level (DL) at Gordion that terminated the Early Phrygian (EP) period
there. I argued that the destruction occurred some time close to 700bc (+/-),
not in the late 9th century bc as maintained in publications and public lec-
tures by the Gordion Team excavators since 2001. The aims of the present
paper are to augment some of the issues I raised previously and to present
additional and relevant information, thus to expand the data available in the
published record. One of the stimuli that generated this review is the grow-
ing number of scholars who have uncritically (to me, without reflection)
accepted the 9th century bc destruction date, and thereby simultaneously
embraced the consequent profound historical and archaeological implica-
tions for first millennium B.C. Aegean and Anatolian archaeology and his-
tory. For example, Prayon (2004: 611) states that the New Chronology has a
weitreichenden Konsequenz fr die historisch-politischen wie auch knst-
lerischen Entwicklungenun Zusammenhnge, which he then proceeds to
document; Prayon and Wittke (2004: 122123) note that das bisherige Bild
der phrygischen Kultur und des Phrygischen Reiches grundlegend vern-
dern . See also Kelp (2004: 286, 293); Strobel (2004: 259, 265268); Genz
(2004: 221, 224); Dusinberre (2005: 4, 10, 220222); Crielaard (2007: 223); and
Summers (2006: 2).
* This chapter originally appeared as Again Gordions Early Phrygian Destruction Date:
ca. 715 +/- B.C., in Studies in Honor of Aykut inaroglu, eds. E. Gen and D. elik (Ankara:
Ekici Form Ofset, 2008), 175187.
602 chapter nineteen
In January 2001, a laboratory (Heidelberg) informed the Gordion Team1

that based on C-14 analysis, the EP citadel of Gordion had been destroyed
ca. 830807/800bc (for details see Muscarella 2003: 225226, 250). The re-
port was immediately and unhesitatingly accepted: for here, was an objec-
tive scientific fact presented by a scientist working in a scientific laboratory,
and thus the previously maintained, for decades, archaeologically argued
and thus subjective dating of the destruction, ca. 700 bc, was rejected. The
New Chronology, as it came to be designated, was declared a fait accompli,
one vigorously upheld by the Gordion Team. The laboratory report was first
publicly announced at the Fifth Anatolian Iron Age Conference in Van in
August 2001 (where, verbally, I first challenged it; for a modified version of
the Van announcement see DeVries et al. 2005). In the same year a brief
statement reporting the New Chronology was published (Manning et al.
2001: 2534; see below). Two years later, a more expanded announcement on
the revised chronology, here provided with brief C-14 specifics (in Table 1)
was published (DeVries et al. 2003); this publication overlapped with Mus-
carella 2003. To date (2007) my 2003 rejection of the presented C-14 date has
been supported in print by only one individual, Keenan (2004)based inde-
pendently on his C-14 interpretation of the analysis (see also Porter 2005: 66
and note 2). The Gordion Team has to date never acknowledged the contro-
versial nature of their new dating.
In Muscarella 2005/2006: 395, note 4, I reported briefly (here expanded)
that I had been informed (in February 2005) that the Gordion Team had
been notified that more recent work on analysis had caused them to lower
the originally presented C-14 chronological range by 40 to 60 years. Such a
change effectively lowers the destruction date, in this calculation, to a time
close to the mid-8th century bc. I made two enquiries about this information
via e-mail to two Gordion Team members (August 2005). One did not answer
my question, except to claim that a work was in progress on Gordion chrono-
logical issues. The other never responded. These unprofessional equivoca-
tions and non-responses tell us that something is wrong within the Gordion
Teamand concomitantly suggest that the information I received in 2005 is
correct. And if correct, then the three-years of silence on the expanded date
range is a serious offense to scholarship.
1 The Gordion Team consists primarily of a quartet, Mary Voigt, Kenneth Sams, Keith
DeVries, and Peter Kunihlom.

again gordions early phrygian destruction date 603
Dendrochronology
The starting point for this dramatic back dating of the destruction at Gor-
dion was generated solely from the C-14 laboratory report. But quite soon
thereafter dendrochronological evidence from structural timbers recovered
in the DL Terrace Building (TB) 2A was soon introduced as both further
and independent evidence supporting the 9th century date (Manning et
al. 2001: 2534). Here four scientists (a Cornell Team) report (in note 28 cit-
ing as the source M. Voigt personal communication), that The last pre-
served ring from construction timbers in Terrace Building 2A is now dated
ca. 883+4/-7bc, and which terminal ring-date approximates construction
of this building, which was destroyed between 830 and 800bc (italics are
mine; see also Voigt 2005: 3031; DeVries 2005: 37 declared with ease that the
destruction occurred not later than 805bc!). The crucial message casually
introduced here (by another university laboratory) is that an unmeasurable
number of missing rings from burnt timbers is irrelevant in chronological
and archaeological investigations; they are not to be questioned by archae-
ologists seeking to determine the absolute date a structure was constructed.
This interpretative leitmotif continued over the years.
Earlier, Voigt (1994: 273) had provided an accurate archaeological analysis
of the absence of chronological value of the very same TB 2A roof beams.
She noted that Kuniholm states (personal communication?) that the
latest ring on these timbers is 908 +/- 37bc (on this specific date see CC3
below). But she correctly saw fit to record that no bark was preserved on the
samples, the number of missing rings is unknown, the beams were burnt
and also suffered excavation-process damage, and, further, that the beams
represent re-used or stored timbers, cut long before this structure was built.
This is a precise and neatly stated archaeological interpretation: but fully
ignored by the Cornell laboratory (in Manning et al. 2001). It should not go
unnoticed that Voigts 1994 disinterested interpretation of the TB 2A timbers
was made before the revelation of the New Chronologyand the later 2001
chronological claim was made after that revelation (see also Kuniholm on
CC 3, below; for others who raise the issue of wood re-use as a fundamental
component in dendro interpretation see James 1991: 323, Keenan 2006: 11,
Mielke 2006: 83, and (yes) Manning et al. 2007: 4). Thus, over the years the
date of the preserved rings went from 908 to 883 +/- bcwith no discussions
regarding the reasons for the changes.
DeVries et al. (2003: 1), and Voigt (2005: 3031) suggested the possibility
of wood re-use for DL structural timbers. Contrary to the two other dates
previously reported for TB 2A908 and 883bc, Voigt now writes that the
latest ring [is] dated ca. 861bc, again with no references to previously
published dates or reasons for the modification.2 The use of Manning et al. in
2001 of TB 2A timbers in support of a specific chronological determination
is therefore puzzling. For aside from the wood-condition problem along
with the correct archaeological conclusions articulated earlier by Voigt, the
new ring determination ignores, indeed contradicts, previously published
dendrochronological evidence and interpretation acquired from another
EP/DL structure, CC3. Wood recovered here manifestly demonstrated that
burnt beams cannot be used to furnish construction dates of structures at
Gordion. I quote Kuniholm (1988: 8), It is clear [sic] that wood was re-
used at Gordion, sometimes many centuries after it was cuta conclusion
he determined from three pieces of wood here that were cut about four
centuries earlier in the 12th century bc, 400 years before eight other wood
pieces from the same structure.3 Concerning the later-cut CC3 timbers,
Kuniholm (1988: 8) informs us that the CC3 Master, composed of eight
samples, [8 separate timbers?] ends in MMTRD 1826sixty two years after
MMT [tumulus MM: Midas Mound]. That is, they were cut 62 years after
the cutting date of the outer logs surrounding MM: MMTRD 1764 represents
the terminal ring that grew just before the tombs logs were felled (ibid: 5).
All the final rings of the tumulus MM logs have not been preserved (even
though the bark is extant) but Kuniholm thinks he has the accurate date.
The cutting date of these logs is presented as 740 + 4/-7bc by Manning et al.
(2001: 2534); previously it had been dated to 718 + 1/-1 bc (see DeVries 2005:
43). To date the ca. 740 bc determination has not been revised (although
given past prouncements it may very well be.4 The 740 MM date is more
or less close to that suggested by some scholars based on artifact analysis
and ancient records Muscarella 1982: 9, idem 2003: 230231, and note 28;
2 Personal communications can be deceptive and misleading and essentially undocu-
mented. For example, a recent email message to me (October 3, 2007) from an interested
party claims that Kuniholm (in a personal communication to him) believes the last ring of
TB2a is 850bc I think it best not to cite it in my text.
3 See James 1991: 323324. See also Muscarella 2003: 227: here I mistakenly thought that
the CC3 wood derived from the post-DL, Middle Phrygian period. In 2003 I was thinking then
only about re-use per se, and therefore missed the chronological implications of the later-
dated timbers.
4 Indeed, one cannot object to modifications of chronology based on more accurate
readings of tree rings, but the changes should be explained, and, each time it should be
indicated that the date is indeed tentative. In one recent example, Kuniholm dated the cutting
year of logs at Ayanis to 655651 bc, causing important and unexpected revisions to royal date
years. Years later he announced that the date was somewhere in the 670s: Manning et al. 2001:
2534, and A. ilingiroglu, in press.
Caner 1983: 9, 201). Missed by me and others, including the Gordion Team, is
that Kuniholm 1988: 67 noted in a chart (but not in the text) that six other
structures in various strata at Gordion contained wood re-used centuries
after the logs cutting date.5
For the record, here are some chronological implications of Kuniholms
1988 report on CC3, working from his designated cutting date for tumulus
MM, and subtracting 62 years. One arrives at these possible dates for the
last preserved ring (not cutting) of wood beams in the DL, 682, 678, or 671bc.
But (not employing Gordion Team dendro analyses), these dates surely only
indicate that EP Gordion was destroyed sometime after these years in the
early 7th century bc (recognized by James 1991: 323324but not by me in
2003).
However, the information and conclusions published about CC3 in
1988 have now been thrown into the bulging file of altered and rejected
data labeled Old Chronology. In an email communication to me (June 29,
2007),6 Kuniholm rejected and modified his previously published conclu-
sions, and changed both the published data and his analysis of the 400-year
tree ring spread along with its chronological implications. He wrote that
the samples were charcoal, (in another email publication, July 12, 2007,
they were horribly burned), and further, that in 1988 he was thennota
beneoperating under the old assumptions about tumulus MM and the
DL, and in retrospect was trying to force a fit where there really wasnt
any (italics mine; this sentence speaks volumes about dendrochronological
analysis and conclusion-formation within the Gordion Team). In addition,
Kuniholm writes that he has since found missing rings that make the
old fit look nothing more than a random resemblance. The result is that
The fit with everything else at Gordion is now good. The end date for CC3 is
MTRD 1595 or 909+4/-7 bc (italics are mine).7 Thus, the previously reported
400-year gap for the CC3 wood has now been officially narrowed; the latest
beams were actually cut (end date) only a century before the New Chronol-
ogy DL date, (and two hundred years before 700 bc). And responding to
my earlier email question about the revised dating of the earlier samples
from CC3, Kuniholm replied (in the same July 12, 2007 communication)
5 And although he mentions timber re-use here, none of the other issues concerning the
wood samples from TB 2Athat they were roof beams, the lack of bark, and missing rings
are discussed.
6 I follow here the Gordion Teams common use in publications of invoking a personal
communication to introduce new, hitherto unpublished, information.

7 Is it a coincidence that the 909bc date is the very same (minus one year) given for TB 2A
aboveor did Voigt or Kuniholm confuse TB 2A with CC3?

that they must be re-used Bronze Age logs. Translation: the older samples
date has not changedthey were centuries-old, re-used timbers from the
Bronze Age, unlike the later beams, now interpreted to be fresh-cut later
all neatly provided by a scientific laboratory. The concepts of missing rings
and re-use are considered meaningful only when they fit into a preconceived
scenario.
I repeat here what I wrote in 2003 (page 227), that re-use of wood at Gor-
dion over long periods of time absolutely precludes any chronological use
of terminal rings as reflecting cutting dates for wood recovered in Gor-
dion structures. The Gordion Team ignores this fact. One of the significant
problems in this matter is that they have unreflectively and unequivocally
accepted, without independent analysis and verification, every conclusion
issued from the Cornell University laboratory. Given Kuniholms (2007) can-
did (and latest) explanation of his ad hoc methodology used in arriving at
his 1988 interpretation, it is appropriate to ask: are scholars now (again) obli-
gated to believe that this revised scientific review does not equally operate
from more recent assumptions about the DL, and that it too is not another
ad hoc attempt to force a fit, this time with the dogma of the New Chronol-
ogy?
In the final analysis archaeologists cannot invoke the currently pro-
claimed 909bc date (last preserved treering) to establish the time of the
destruction of CC3 and the DLbecause of the possibility of wood re-use,
and because the latest preserved ring says nothing about the date of con-
structing a building.8 Indeed, although preached and widely accepted oth-
erwise, it is a false dichotomy to claim that dendrochronology is an objective
science to be privileged over subjective archaeology. It is therefore signif-
icant (and surprising) to find this correct position precisely stated recently
by the Cornell Laboratory stafffor the first time: Dendrochronology is not
an exact science (Manning 2007: 3).
Until the published (and unpublished) reports cited above are reinvesti-
gated and published in detail, giving all information related to the condi-
tion of the wood involved, the presence or absence of bark, the evidence
and reasons for both the original and subsequent date changes, and the
8 Concerning Kuniholms continuous modifications of and changes to data and conclu-
sions over the years regarding C-14/dendrochronological datingoften without reference

to previous claims, see Mielke 2006: 8283 and note 17, 90, with references to missing rings;
Keenan 2006: 5. James (2002) records the series of dates previously published by Kuniholm
for the tumulus MM burial: 547, 757, 718 bc; he also critiques the present date of 740 bc. For
the practice of wood re-use in Anatolia from earliest times see M. and A. zdogan 1998: 589.
methodologies employed, including recognizing re-use of wood, no archae-

ologist can use the TB 2A and CC3 timbers for any chronological determina-
tionsmost certainly not relating to the DL date (see also below). All the
Cornell laboratorys claims remain in abeyanceas does also my tentative
7th century dating (above) based solely on the 1988 report. My reading of
Manning et al. 2007: 1, 5 is that the Cornell laboratory slightly revises past
practice and in the future will review and critique (reanalysis) the reports
its staff has published, and will conduct a re-study of all the published Gor-
dion wood reports.9
Strobel (2004: 266267, 271, and 275, note 4) with ease dismisses my argu-
ments against the Gordion Teams use of dendrochronology. As the Fllda-
tum [sic] der Hlzer for TB 2A he proclaims it to be 861/883bc, and for CC3
it is 912919bc; missing rings, wood re-use, or revisions of chronologies over
time are to him mere minor, pedantic issues of no archaeological concern.
Genz (2004: 221, 226) cites the dendrochronological together with the C-14
data as evidence for the 9th century dating of the DL at Gordion.
In 2004 Kuniholm (page 1) accelerated the use of dendrochronology to
affirm the New Chronology, here again invoking the last-ring solecism by
citing the rings of four logs (Voigt in Yildirim and Gates 2007: 31 says it is one
log) recovered from the base of the Early Phrygian fortification wall as data
that reinforces the New Chronology. Because on these logs the last existing
ring (no bark preserved) is 862 bc, they are therefore taken to indicate that
the EP settlement and its fortification wall appear to be a ninth-century
affair (italics are mine). Not mentioned is the significant different conclu-
sion he forcibly presented in 1988, or the conclusions of Voigt in 1994 and
2005, or his ignoring the problem of using bark-less wood for absolute dat-
ing (see Manning et al. 2007: 5), and also the possibility of re-use of earlier
wood. Voigt (in Yildirim and Gates 2007: 31), ignoring her earlier cautious
stance, also cites the walls log as evidence to support the New Chronology
another example of the Gordion Team automatically accepting a scientific
decree.10 All I discuss here has now been manifestly confirmed, for recent
9 I came upon Manning 2007 after I had essentially finished writing this paper, courtesy
of D. Keenan. Aside from informing us (subtlyI had to read it several times to understand
it) of the review to be undertaken, the report provides a clear discussion about dendro
methodology and the many problems involved in analyses: see also Mielke 2006: 7784.
10 It is relevant here to remind scholars of the (if correct!) spectacular discovery by
Kuniholm of a central joist within a modern house on the Black Sea that was cut 6200 years
ago (Aegean Dendrochronology Project December 2001 Progress Report: 8). Had this house been
burned down 50 years ago and recently investigated, no doubt some dendrochronologists
would have dated the house to the early Neolithic period.
dendro examination at Gordion has proved (yet again!) that timbers can-
not be used as chronological markers. In the most recent Cornell University
Weiner Laboratory report (December 2007: 3) is the statement by Sturt Man-
ning that: The date for the last preserved ring of juniper logs from Building
A, from the MP (i.e. post DL!) citadel, is 991 + 4/-7bc. But unfortunately his
report refrained from confronting the fact that the MP period is dated by
the Gordion Team post DL, to the late 9th early 8th century bc (but to me
late 8th at the earliest, or the early 7th century bc), and that either way we
have here evidence of reused beams; and that a beam with its last ring dat-
ing almost a century earlier than those from the EP period buildings TB 2A
and CC3 (above), is not employed here using the same methodology to date
the MP period to the 10th century bc!
Dendrochronological dating of tumuli has suffered the same blinkered
mis-treatment: in several instances we are commanded to accept as an abso-
lute chronological determinant the last preserved ring of a tombs timber
construction. Thus Strobel (2004: 271, 276 and note 10) reveals to archaeolo-
gists that the plundered and destroyed Kayran Mevkii tumulus (a Gordion
tumulus) was constructed in the 9th century bc (862 + 7/-3)because of a
(not fully published) dendrochronological dating of burnt logswith no
bark; and no tomb artifacts were preserved. In the same sentence (p. 276)
we are also commanded to know that there exists other Phrygian tumuli
constructed in the 9th century B.C, tumulus W (see below), and also the, to
me, 8th century bc Mamaderesi tumulus (another Gordion tumulus), which
date is known to him auf die Funde; in a footnote (10) he mentions the 80
fibulae recovered. The chronology of the tumuli as proclaimed by Strobel,
is of major importance to his general Early Chronology project, and he sim-
ply announces that the 80 fibulae were made in the 9th century bc (pace
Muscarella 2003: 231). This chronology can be based only on an a priori deter-
mination about the DL (and consequent tumulus construction) chronology:
i.e., a circular argument. The Mamaderesi fibulae include Types XII, 2, 9,
13, 13A, 14, and 14A, only one of which type occurs in tumulus W (9th cen-
tury to Strobel), and there are no 7A examples here, although they are quite
common in W. A number of the Mamaderesi fibula types were recovered in
manifestly 8th century bc tumuli, such as (even to Strobel) MM, which con-
tained XII, 7, 9, 11, 13, and 14 examples. The late 8th century or slightly later
tumulus S-1 (viz. DeVries 2005: 3940, notes 46, 43; Kohler 1995: 115140)
also has two XII, 14A buckles (ibid. 127128, pl. 66), a late form (compare
Strobel 2004: 276 and note 10). An appropriate question raises itself based
on artifact analysis: why should archaeologists not assert that tumulus MM
must also be dated to the 9th century bc, inasmuch as it shares with the
DL and Mamaderesi tumulus XII, 9, 13, and 14 fibulae forms? (See also the
discussion of the clay foundation deposit below.)
For the record: that tumulus MM did not contain the burial of king
Midasbecause of its construction date determined by dendro (more or
less) and artifact analysisand that it belonged to an earlier ruler, probably
Midas father, has been argued for a long time: by R.S. Young after its exca-
vation, and subsequently by a good number of other scholars, including me,
for decades (see Muscarella 1982: 9, idem 1995: 99, note 13, idem 2003: 231,
note 28). This has been ignored by the Gordion Team in lectures and writing:
viz. Kohler 1995: 192, 228; Manning et al. 2002: 2534; DeVries 2005: 42; DeVries
et al. 2003: 2; DeVries et al. 2005: 45. Voigt (in Yildirim and Gates 2007: 311)
writes as if we now know for the first time ever that Midas is not buried in
MMT. And Strobel (2004: 276) indicts me for continuing to maintain there
is a relationship between MM and King Midas and his family. This specious
charge requires no response, but I will state simply what has often been rec-
ognized as obvious: I believe a long with other archaeologists, that MM is
manifestly a royal burial based on its labor intensive size and construction,
its locus, and its artifact deposition and chronology coincide with/are quite
close to that of the historical King Midas/Mita. Its chronology strongly sug-
gests however, that Midas himself is not buried there.
The Gordion Team (DeVries et al. 2005: 46) asserts its position loud and
clear, The radiocarbon and dendrochronological evidence provides a firm
and consistent absolute chronology for ninth [Early Phrygian] and eighth
century [Middle Phrygian] Gordion (italics are mine). (What is consistent
is their circular-citation of each others publicationsbut very few of other,
relevant publications.) For analytic criticisms and rejections of the scien-
tific consistent claims blithely made by DeVries et al. 2005, see Keenan
2004, idem 2006, James 1991: 322, idem 2002. Artifacts are not mentioned
as evidence in the above quoted claim, but are mentioned in the article
itself (p. 45), with the claim that they independently indicate a late 9th cen-
tury New Chronology date; see also DeVries et al. 2003: 2, Voigt 2005: 31, and
idem, in Yildirim and Gates 2007: 311. These artifacts include fibulae, horse
trappings, vessel attachments, ivory horse pieces, attachments, pottery, and
orthostates, all of which are chronologically evaluated in Muscarella 2003:
237245, although not confronted by the Gordion Team. On an (attempted)
objective basis, i.e. not a priori using the old date of ca. 700bc for the DL
as a guide, I argued that the DL artifacts are 8th, not 9th century produc-
tions. I have not changed my mind, and those who disagree should confront
each of my evaluations. Strobel (2004: 278), consistent in asserting, but not
demonstrating a 9th century date of each DL artifact, does not.
A viable argument in the present framework would be to state up front

that consequent to any scholars acceptance of the late 9th century C-14
chronology for the DL, it therefore must follow that its artifacts are dated to
that time. This has been accomplished formally by Prayon and Wittke (2004:
123), who date one of the ivories from the DL clearly aufgrund ihrer Fundsit-
uation nach der neueren Chronologie wohl noch in das 9. Jh. zu datieren ;
also Prayon 2004: 614; Genz (2004: 225226) says we must now date certain
types of central Anatolian pottery and fibulae to the 9th century bc because
of the revised Gordion chronology. If one accepts the C-14 New Chronology,
these will be seen as correct and necessary assessments: although the evi-
dence actually indicates otherwise.
I focus now on a summary of the fibulae recovered in the DL, and add
some scholarly commentary. I also add an amplification concerning the loci
at Gordion of fibulae and another artifact, socketed arrowheads. The latter
has not hitherto been discussed, but is a major component of evidence that
must be used to determine the date of the DL at Gordion.
Fibulae
The Phrygian fibulae, commonly known as Type XII, recovered in the DL

are dated both by comparison with those from the tumulus burials and from
outside comparanda, and provide a general chronology covering the second
half of the 8th century for the DL examples.11
As I reported in 2003 (229234), the DL fibulae types include two Type XII,
5; 22 XII, 7A; one XII, 9; one XII, 13 or 14; and one XII, 14, total of 27 Phrygian
fibulae. The DL also contained 14 imported foreign Aegean and Near East-
ern fibulae, dating to the 8th7th century bc and therefore relevant to DL
chronology investigations (Muscarella 2004: 233236), pace Voigt (2005: 31),
who states that no Greek material had ever been found in the Early Phry-
gian Destruction Level. The Aegean fibulae parallels are not sehr vage as
Strobel asserts (2004: 276; Maya Vassileva will eventually publish them in
detail). Aegean fibulae, along with socketed arrowheads, were also recov-
ered in the fill over tumulus B (Kohler 1995: 2122, pl. 11). These arrowheads
demonstrate a late 8th or 7th century date for the construction of the tumu-
lus, as well as a nomadic peoples presence at Gordon (at some undeter-
mined specific time; see below).
11 I use standard terminology here and not Caners 1983 unnecessary new terminology,
intended to replace Blinkenberg, Muscarella, Boehmer, etc., and pace Strobel 2004: 276,
note 5.
The Phrygian fibulae from several of the tumuli parallel the DL fibulae.
The most common type from the burials, Type XII, 7A, occurs in tumulus W,
KIII, KIV, S and G; XII, 5 in KIII; XII, 9 in KIII, KIV, MM, S-1, and Mamaderesi;
and Type XII, 14 in KIV, MM, S-1, and Mamaderesi. In the Ankara tumulus
excavated by Sevim Bulu, Type XII, 7A, 9 and 14 forms also occur together
(Muscarella 2003: 233). All these tumuli are dated to the second half of the
8th century. A review of earlier scholars determinations on the relative
construction dates of the tumuli yields the conclusions that MM, P, KIII,
and KIV are relatively chronologically close to each other, W is earlier, and
all date to the 8th century bc. DeVries (2005: 43) dates tumulus MM before
Mamaderesi. Strobel (2004: 267, 276277) dates tumulus KIII and P close to
the time of the DL, and for tumulus W makes another throwaway assertion
of a date: bestimmte Fundstcke belegen ein Datum im spten 9. Jh. V. Chr.,
but which Fundstcke are not shared with us. Wittke (2004: 257, note 361;
258, 268) dates tumulus W to 850bc and KIII to the first half of the 9th
century bc. Crielaard (2007: 223) dates W to ca. 850 bc on the basis of recent
radiocarbon dating, which in fact does not exist for this tomb (although he
may have meant the recent DL carbon dating). All these chronologies follow
manifestly from the authors acceptance of the New Chronologyfor not
one provides detailed artifact analyses. Tumulus W is indeed the earliest,
in my opinion dated ca. 750 +/- bc (Muscarella 1982: 8, idem 1995: 97, and
idem 2003: 229; also Kohler 1995: 191192).
Post 2001 the Gordion Team also supports a 9th century bc date for
tumulus W (verbal communications) primarily because of the presence
therein of XII, 7A fibulae, which to them are chronologically paralleled by
their presence in the DLand therefore 9th century bc. If this is the case,
then KIII, KIV, and G are also to be dated to the 9th century. But what of the
other later DL fibula forms, Types XII, 7, 9, and 14, all of which occur in other
tumuli, especially MM, which tumulus all the involved scholars date more
or less to ca. 740 bc?
For a discussion of the chronological relationship of the Gordion tumuli
based on artifact analyses, the comparisons among them, and conclusions
that they were not constructed far apart in time one another within the
2nd half of the 8th century bc, see Muscarella 1982: 8, idem 1989: 337342,
and idem 2003: 229230. Whether these conclusions are correct or not, the
analyses were based on an analysis of the finds and not on a preconceived
date for the DL. Strobel (2004: 276277) disagrees with this chronology, but
without meaningful discussion of the parallels brought forth. In the final
analysis, scholars will have to decide which chronological interpretation in
fact drehen sich immer wieder im Kreise. The vital point that I repeat here
is that all the Phrygian fibulae forms recovered in the DL are also present in
the tumuli: that are accepted by all parties to have been constructed in the
8th century bc. Moreover, no Phrygian fibulae recovered in the West (none
are Type XII, 7A) have been dated by pre-8th century bc (Muscarella 1989:
338339, and note 21).
Numerous other artifacts recovered in the DL have yet to be published,
and thus remain to be analyzed for contributions to the chronological issues
(Maya Vassileva is presently engaged in publishing the metal remains).
Clay Foundation Deposit
Another locus at Gordion for the fibulae remains to be evaluated, one that
reinforces the Old Chronology. Only recently did I become aware that there
are more fibulae available for chronological discussions than I had hitherto
realized, that for some inexplicable reason I had missed. I argue that these
fibulae must be brought into the discussion about the date of the DL, and
that they reinforce my original artifact analyses and dating.
Mary Voigts excavations (Voigt and Henrickson 2000: 52) produced a sig-
nificant fundamental adjustment in our knowledge and understanding of
the nature of the cultural and chronological sequence at Gordion follow-
ing the destruction. She determined that the citadel was resettled above the
DL, not 150 years later (as posited by Rodney Young), but quite soon after,
with little or no gap (Muscarella 2003: 227228). In DeVries et al. 2003: 2 (for
which Voigt was a co-author) the rebuilding time was modified to a gener-
ation after the destruction. Later, however, Voigt (2005: 32) reinforced her
original conclusion (ignoring her 2003 position), claiming that the recon-
struction began immediately after the EP destruction.
If this stratigraphical/chronological adjustment holds up it seems the
labor-intensive and well-planned rebuilding process began soon after the
destruction of the EP citadel, although whether within a year, or a few years,
or a decade (more or less) remains unknown. This well-organized rebuild-
ing was accomplished first by leveling the DL debris and some walls, then
covering most of the DL with a massive clay deposit two meters thick, within
which were set the stone rubble foundations for the structures forming the
settlement that was built directly on top of the clay (see Voigt 1994: 273, pls.
25.5, 6.2; 2005: 33, fig. 37A, B). Closely copying the plan of the EP citadel
which is another indication that rebuilding occurred soon after the destruc-
tion (Voigt 2005: figs. 3-3, 3-4)the post-clay settlement is designated Mid-
dle Phrygian.
It is worth considering here, at least parenthetically, whether this massive

clay deposition was solely an engineering/architectural necessity, to cover
the DL with a thick foundation in order to rebuild securely the settlement
above. But one may legitimately wonder if the clay deposit may also have
been a ritual burial of the destroyed city of King Midas (who committed
suicide when the city was destroyed!), or a burial in honor of the Phry-
gian/Gordion Goddess Matar, a burial directly linked to the rebuildingon
the same plan as the (sacred?) buried city below. There is a tradition of the
ritual burial of sacred structures in Anatolia going back to the pre-pottery
Neolithic period, as neatly documented by zdogan and zdogan (1998:
589592; see also Schmidt 2001: 46). As the zdogans noted, the leveling of
buried structures and preservation of what remains of walls are followed by
a rebuilding immediately over the earlier structure (for a deliberate cov-
ering deposit at Nuzi see Bjorkman 1999: 106). That is, events precisely as
happened at Gordion. Bjorkman (1999) has cited later, historical examples
of ritual burials of sacred structures at a number of Mesopotamian sites. A
still later example of such burial is that of Nush-i Jan, a 7th century temple
site in Iran (Muscarella 1988: 207, 208, note 2). I suggest that the possibility
of ritual burial be considered as a viable interpretation for Gordions transi-
tion from EP to MPeven if it cannot be demonstrated as historical reality
at present.
A number of artifacts were recovered from the clay foundation; the clay
was clean but not quite sterile (as claimed by Strobel 2004: 266267). Some
clay-derived artifacts clearly date from a much earlier period (Dusinberre
2005: 10); I excavated there a beautifully hand-cut, apparently Neolithic,
obsidian blade (not published). Other finds include Phrygian terracotta
and metal artifacts. Those of us digging at the site believed that these
artifacts had been lost or discarded, then accidentally picked up in the
buckets by Middle Phrygian workmen from the clay/mud sources (that
there were at least two sources is revealed by deposits of different colors,
orange and gray). The Phrygians involved in the bucket brigade may also
have dropped some contemporary objects (fibulae?) accidentally. Artifacts
from the clay could also have derived from the DL debris, displaced in the
leveling process (on this see M. and A. zdogan 1998: 589). A significant
example was a broken stone bearing an inscription recovered from one of
the MP structure foundations built into the clay (Muscarella 2003: 246). This
stone most probably derived from the DL itself, picked up along with many
other available EP wall stones that were obviously in full view and readily
available to the re-builders, or it may have come from an off-site source.
In either event, the stones placement in the foundations set in the clay
Figure 1.
occurred (according to all scholars concerned) soon after the destruction.

And, as it was set into the foundations, it predates the construction of the
MP structures.
Among the fibulae recovered in the clay deposit were one Type XII, 7,
two XII, 9, three XII, 14 forms (Figure 1, for one example: B 1304; hitherto
unpublished). It is possible that some of these are earlier than the DL, picked
up in the clay sources (see Caner 1982: 2, note 15 for this source recognized
in tumuli formation fill), but it is worth noting that all have parallels with
those deposited in the late 8th century bc tumuli, MM, S-1, and Mamaderesi
being the outstanding examples. Types XII, 9 and 14 also occur in the DL.
And Types XII, 7 and 9 are also paralleled on reliefs at Khorsabad and Ivriz,
both of which document a late 8th century bc floruit for these formsthe
XII, 9 at Ivriz has a double pin mechanism and lock-plate, exactly as found
on examples from tumulus MM (Muscarella 1982: 78; Caner 1983: 9, 173, 201,
Taf, 62). Strobel (2004: 277, note 15) casually (and unsuccessfully) dismisses
the chronological value of the Ivriz relief. Found in the MP South Cellar were
fibulae of Types XII, 13, 9, and 14, the latter two surely relatively late forms,
and paralleling those from tumulus S-1, later than those from tumulus MM.12
12 DeVries 2005 includes a modern reconstruction of the South Cellars stratigraphy (the
section, Figure 42, was made in Philadelphia, not Gordion), and it presents confusing dating
and loci of the Greek pottery excavated there. It is disturbing that DeVries made no mention
in this article of his and K. Sams claims made years earlier about the loci and associated
Figure 2.
Arrowheads
Another group of artifacts recovered in the clay include bronze socketed

arrowheads (hitherto unpublished). When I recorded the corpus of socketed
arrowheads excavated from various geographical areas in the Near East
(Derin and Muscarella 2001: 195, and Muscarella 2006: 156) I reported that
no such arrows could be attributed to the destruction level at Gordion.
Inexplicably, I was not then thinking of reviewing the evidence from the
clayan extension of the DL; that thought occurred to me later, hence the
present review.
One example (B1196; Figure 2, left) was recovered by Rodney Young in
the clay and rubble wall beds of Bldg M dug right down to the top of the
burned fill overlying Megaron 3 From the clay: 5769 B1175. Bronze wing
problems relating to the Greek sherds under discussion (as if for the first time). For in fact, in
earlier publications, DeVries claimed that the pottery derived from pre-destruction contexts
although none is from a stratigraphically instructive context, and Sams claimed that the
fragments had been recovered scattered and dislocated within the city and elsewhere! I
discussed these issues in Muscarella 2003:241243, with references. It is not cited by DeVries.
For chronological problems at Bogazky regarding imported Greek pottery, see Muscarella
2005/2006: 394. I suggest that we have not heard the last about the South Cellar stratigraphy
or of the loci of the Greek sherds.
with fragments of gold foil[;] 5903 B1196: bronze arrow point (R.S. Young
Field Book, Vol. 78: 14). B1196 is a socketed trilobate arrowhead. Another
(7876 B1503; Figure 2, right) was found below a white floor covering the
clay: work continued digging clay. Beneath the white floor was found
a well-preserved bronze arrowhead, c 4cm long. 3 flanged i.e. a trilobate.
(R.S. Young Field Book, Vol. 114: 5051).
There is a third example to be considered (89 SF# 16 Operation 11: a bilo-
bate). This arrowhead was discovered in 1989 by M. Voigt, who generously
gave me the following information (verbal 1992, and in recent email com-
munication): The arrowhead from OP 11 came from a drain that is part of
the construction of YHSS 6B or (early) Phrygian The stub [on which the
arrow was recovered] is under a drain that was built as part of YHSS 6A, the
construction phase that was mostly standing when the fire [DL] took place
The problem is that RSY dug down to the level of the PAP/6B and then
below it to the wall stub in 1963 and both lay exposed till we cleared it in
1989 a lot of people went right across that area for many years of exca-
vation so you cant exclude the possibility that the arrowhead was dropped
there [post 1963]. We found an Attic black figured sherd on the exposed sur-
face [which?] as well. Since the arrowhead was tucked between stones, it
could be in situ but if it is a critical part of any argument, the possibility
of it being dropped cannot be excluded. Indeed, I accept this conclusion
as objectively viable. I also add that the arrow could have been dumped
along with the fibulae and the two arrowheads mentioned abovewith the
clay deposit. This arrowhead is best left in abeyanceneither included nor
excluded in the discussion. I record it as evidence of a possible third example
of a nomadic arrowhead from the DL.
The interesting and most important fact about the arrowheads found in
the clay is this: not a single excavated bronze socketed arrowhead is known
from excavations anywhere in the Near East that pre-dates the 7th century bc.
Some surely may have been present since the very late 8th century bc, but
up to the present none this early has been identified. This chronological
situation has been fully documented by Derin and Muscarella 2001: 197, pas-
sim, and Muscarella 2006: 157158. Thus, the arrowheads in the clay deposit
manifestly document a post-late 8th century bc date for the DL. Indeed, sock-
eted arrowheads traditionally have been culturally associated with intrusive
nomadic peoples (first recorded in the late 8th century bcthe earliest date
is 715 bc, for the Cimmerians) but they were soon thereafter employed by
other cultures (Derin and Muscarella 2001: 197203; Muscarella 2006: 157
158). However, their presence at Gordion in a post-late 8th century bc clay
deposit context surely suggests a Cimmerian presence, as I and others have
argued for some time, pace the Gordion Teams rejection (Muscarella 2003:
247249; Muscarella 2005/2006: 395. For possible Cimmerian presence at
Norsuntepe and Bogazky, both because of artifacts recovered including
socketed arrowheads. And in northern geographical locations, see Derin and
Muscarella 2001: 199200; for a probable nomadic burial at Gordion in tumu-
lus KY, where two horses were buried, but no arrows (or Phrygian fibulae!)
were recovered, see page 195).
Fibulae do not lie (fib); to this (growing) list of truth speakers we now add
bronze socketed arrowheads. Both occur within the clay deposit/platform
overlying the DL, constructed soon after the destruction and almost simul-
taneously with the erection of the MP settlement. Both the fibulae and the
arrowheads date the DL to ca. 700 +/- bc, a time just preceding the beginning
of the MP period. The New Chronology decreed for the DL must therefore be
rejected because: the C-14 data have been accepted ad hoc without confir-
mation from a second laboratorys analysis and discussion of statistics (see
also above); the offered dendrochronological conclusions and their inter-
pretations are flawed; and archaeological analysis of the artifacts preserved
in the DL and the clay do not support the New Chronology, they contravene
it. Fibulae may indeed be cited as an axiomatisches Argument, they most
certainly are one of archaeologys best chronological indicators, a classic
Leitfossil, as are socketed arrowheads. Such realia cannot casually be ignored
or manipulated.
Bibliography
J. Bjorkman 1999: How to Bury a Temple , in Studies in the Civilizations of Nuzi

and the Hurrians, vol. 10, ed. D.I. Owen and G. Wilhelm: 103122.
E. Caner 1983: Fibeln in Anatolia, Mnchen.
J.P. Crielaard 2007: review of Wittke 2004, in Bibliotheca Orientalis LXIV, 1/2: 220226.
Z. Derin and O. White Muscarella 2001: Iron and Bronze Arrows, in A. ilingiroglu
and M. Salvini, Ayanis 1, Istituto per gli Studi Micenei ed Egeo-Anataolici, Rome
K. DeVries 2005: Greek Pottery and Gordion Chronology, in Lisa Kealhofer ed., The
Archaeology of Midas and the Phrygians Recent Work at Gordion, Philadelphia:
3755.
K. DeVries, P. Kuniholm, G.K. Sams, M.M. Voigt June 2003: New dates for Iron Age
Gordion, Antiquity 77 (Internet: http://antiquity.acuk/ProjGall/devries.html).
K. DeVries, G.K. Sams. M.M. Voigt 2005: Gordion redating, in A. ilingiroglu and
G. Darbeyshire, eds., Anatolian Iron Ages 5: 4546.
E. Dusinberre 2005: Gordion Seals and Sealin, University of Pennsylvania Museum.
H. Genz 2004: Erste Anstze zu einer Chronologie der frhen Eisenzeit in Zen-
tralanatolien, in Die Aussenwirkung des spthethitischen Kulturraumes, M. No-
vk et al., eds. Ugarit-Verlag: 219228.
P. James 1991: Centuries of Darkness, London.

2002, The Dendrochronological Debate, Minerva XIII, 4: 18.
D.J. Keenan 2004: Radiocarbon Dates from Iron Age Gordion are Confounded,
Ancient West and East, 31, 1: 100103.
: Anatolian tree-ring studies are untrustworthy, Informath (Internet): 120.
U. Kelp 2004: Der Einfluss des spthethitischen Kulturraumes auf Orthostaten in
Gordion, in Die Aussenwirkung des spthethitischen Kulturraumes, M. Novk et
al., eds. Ugarit-Verlag: 284295.
E. Kohler 1995: The Lesser Phrygian Tumuli, Part I The Inhumations, University
Museum, The University of Pennsylvania.
P. Kuniholm 1988: Dendrochronology and radiocarbon Dates for Gordion and
Other Phrygian Sites, SOURCE, Notes in the History of Art V11, 3/4: 58.
2004: Aegean Dendrochronology Project December 2004 Progress Report,
Cornell University.
S.W. Manning, B. Kromer, P. Kuniholm, M.W. Newton 2001: Anatolian Tree Rings
and a New Chronology for the East Mediterranean Bronze-Ages, Science 294:
25322535.
S.W. Manning, C. Griggs, Maryanne Newton, Tomasz Wazny 2007: Summary and
Discussion of Procedures at the Malcolm and Carolyn Weiner Laboratory for
Aegean and Near Eastern Dendrochronology, Cornell Tree-Ring Laboratory, In-
ternet: www.arts.cornell.edu/dendro/summary%20of%20procedures.pdf.
Mielke, Dirk Paul 2006: Dendrochronologie und hethitische Archologieeinige
kritische Anmerkungen, BYZANS 4: 7794.
O. White Muscarella 1982: King Midas Tumulus at Gordion, The Quarterly Review
of Archaeology, December: 710.
1989: King Midas of Phrygia and the Greeks, in Anatolia and the Ancient Near
East, Studies in Honor of Tashin g, K. Emre et al. eds., Ankara: 333344.
1995: The Iron Age Background to the Formation of the Phrygian State,
Bulletin of the American Schools of Oriental Research 229/300:91101.
2003: The Date of the Destruction of the early Phrygian Period at Gordion,
Ancient West and East 212: 225252.
2005/6: review of H. Genz, Bogazky-Hattusa XXI: Byykaya I. Die Keramik
der Eisenzeit , in Archiv fr Orientforschung 51: 393396.
2006: Bronze Socketed Arrowheads and Ethnic Attribution, in The Golden
deer of Eurasia, J. Aruz. et al., eds. Metropolitan Museum of Art: 154159.
M. and A. zdogan: 1998: Buildings and the Cult of Buildings, in Light on Top of the
Black Hill, Studies presented to Halet ambel, Istanbul: 581593.
R. Porter 2005, Carbon Dating, The Situation in 2005, Journal of the Ancient Chro-
nology Forum: 6366.
F. Prayon 2004: Zum Problem von Kultsrten und Kultbildern der Anatolische
Muttergttin in 8. Jh. V. Chr., in Anadoluda Dogu, H. Iskan and G. Isin eds.,
Istanbul 2004: 611618.
F. Prayon and A.M. Wittke 2004: Die Aussenwirkung des spthethitischen Kultur-
raumes auf Zentral- und Westanatolien und Zpperen, in Die Aussenwirkung des
spthethitischen Kulturraumes, M. Novk et al., eds. Ugarit-Verlag: 121126.
K. Schmidt 2000: Gbelki Tepe, Southeastern Turkey A Preliminary Report on the
19951999 Excavations, Palorient 26: 4554.
K. Strobel 2004: Neue Fragen zur Chronologie Gordions und Anatoliens im 1.

Jahrtausend v. Chr., in Die Aussenwirkung des spthethitischen Kulturraumes,
M. Novk et al., eds. Ugarit-Verlag, 2004: 259284.
G. Summers 2006: review of Dusinberre 2005 in Bryn Mawr Classical Review, 03.27:
15 (Internet).
M.M. Voigt 1994: Excavations at Gordion 19881989, in Anatolian Iron Ages 3,
A. ilingiroglu and D.H. French eds., British Institute of Archaeology, Ankara:
265293.
2005: Old Problems and New Solutions, Recent excavations at Gordion in
Lisa Kealhofer ed., The Archaeology of Midas and the Phrygians Recent Work at
Gordion, Philadelphia: 2235.
M. Voigt and R. Henrickson 2000: Formation of the Phrygian State: the Early Iron
Age at Gordion. Anatolian Studies 50: 3754.
Anne-Maria Wittke 2004: Musker und Phryger, Wiesbaden.
B. Yildirim and M.H. Gates 2007: Archaeology in Turkey: 20042005, American
Journal of Archaeology 111: 275356.
chapter twenty
URARTIAN METAL ARTIFACTS: AN ARCHAEOLOGICAL REVIEW *
Abstract: Metal artifacts constitute the main component of the cultural material
known from the ancient state of Urartu. The great majority of these artifacts avail-
able to archaeologists for cultural analysis derives from plundered sites, which
proveniences, whether from sites in modern Turkey, Armenia, or northwestern Iran,
are unknown. In fact, only a minority was obtained from excavations at known sites
and tombs. At the same time there is a large quantity of publications on Urartian
metal artifacts, most of which are in museums and private collections in Europe,
the United States, and Japan. Most of the publications, many by museum person-
nel, ignore the reality of the lack of provenience and present subjective conclusions
regarding provenience and interpretation. This paper discusses the consequences
of the non-archaeological methodology of these publications, the nature of which
many archaeologists remain unaware, and how this behavior has interfered with
achieving an accurate knowledge of Urartian culture.
Scholars attempting a cultural study of Urartian metal artifacts experience
immediately several problems. First is acquiring the large number of publi-
cations spread over many venues. Then when investigated, one soon recog-
nizes that the majority of the material published has no archaeological prove-
nience, rather they have a large number of museum, dealer and collector
provenances in many nations. These non-archaeological provenances, a cul-
tural artifact of modern consumerism and greed, signify that only a limited
corpus of archaeologically documented data are available for meaningful
study, juxtaposed to a larger corpus that does not possess this quality. And
further, one soon becomes aware that, notwithstanding the fundamental
significance of the provenance/provenience, locus/non-locus relative pro-
portion reality, this issue remains basically unperceived, certainly under-
appreciated, and at times rejected in scholarly discussions. Only a very few
scholars call attention to it. Others note the disproportionate nature of the
excavated data merely as an aside, and then proceed to argue an archaeo-
logical interpretation that ignores the ontological backgrounds of the dis-
tinct corpuses. The consequence is that an entrenched subjective, non-
anchored, anything-goes publication pattern has come to prevail, and its
* This article originally appeared as Urartian Metal Artifacts: An Archaeological Review,
Ancient Civilizations from Scythia to Siberia 12, 1/2 (2006): 147177.

622 chapter twenty
success, promoted as normal archaeological discourse, has become a model

for others to follow. This pattern has led to intellectual disorder, a serious
dysfunction in studies of Urartian metal artifacts, which are in fact a major
component of the cultures artifact productions.
How have scholars characterized and understood the modern history of
the extant corpus of Urartian metal objects in their attempts to form cultural
and historical conclusions? How has it been determined which objects in
the corpus are manifest archaeological data, which are limited, and which
are demonstrably of limited or non-existent evidence in seeking archaeolog-
ical truth? The aim of this paper is to demonstrate that many Urartian schol-
ars have not distinguished plundered/purchased objects from excavation-
generated artifacts, and that they have accepted as proper research strategy
the penetration of the dealer/museum/curator-generated publications into
the archaeological realm.
A review of the majority of the relevant publications reveals ignorance
of or disinterest with the modern nature of the unexcavated, plundered,
objects presented, fulsome acceptance of the often cited alleged/said to be
found at X, angeblich archaeological proveniences, and formed subsequent
conclusions based on them. Concisely stated, for various cultural, political,
and personal reasons the majority of scholars ignore or are utterly unaware
of the vital role played by museum curators and directors in the forma-
tion of the dysfunction that prevails in Urartian archaeological scholarship,
that museum agendas are often antithetical to, certainly rarely allied with,
archaeological goals. Let it be clear up front, one cannot challenge the use
of unexcavated artifacts as data in archaeological research: for the obvious
reason that these orphaned objects exist as realia in the modern world,
they have an ancient past: pace our concern for the destruction generating
their existence in the modern world, and how they came to be exhibited
in rich plunderers (a.k.a. museums and collectors) vitrines. The issue to
be addressed directly is not that archaeologists can or should ignore these
unexcavated strays, but rather in what manner and within what parame-
ters archaeologists seeking archaeological knowledge should or should not
employ these unanchored phenomena with inherently limited information.
As much a challenge as a reality, scholars beginning research on Urartian
metal artifacts must immediately understand that the majority of Urartian
objects to be investigated are unexcavated.
O. Belli claimed (15 years ago) that about ten thousand Urartian objects
had been plundered from eastern Turkey in the previous twenty years.1
1 Merhav 1991, 44.

urartian metal artifacts: an archaeological review 623
Whether this is a correct number can be debatedbut in the years since,

many more have been plundered. And the photographs of but two large
groups of plundered material published by H.-J. Kellner suggest the num-
ber is large: one illustrates hundreds of Urartian bronze and iron artifacts
massed together on a table in the cellar of a single Kunstgalerie,2 a collec-
tion complacently described by Kellner in fluent museum-speak as einen
seltenen Glcksfall, the other a mass of helmets. The photographs demon-
strate the reality that a photo is worth a thousand words: the plundered
number is huge, and one of the provenances for Urartian metal artifacts
is Munich, Germany. While some scholars recognize the problem (some,
vaguely), too large a number conduct research and publish while wearing
horse blinkers (a common Urartian artifact, some of which have been exca-
vated); they accept the command ordered in many publications: all objects
designated here to be Urartian and assigned to a specific site locus have an
equal-opportunity cultural value equal to excavated objects; both sets must
be accorded equal epistemological value as bona fide archaeological data.
It is appropriate to initiate an archaeological investigation of the pub-
lication of Urartian metal artifacts with reference to an often-cited com-
prehensive, important but archaeologically deceptive and flawed catalogue
that succinctly serves (albeit unintentionally) as a paradigm of the present
state of Urartian scholarship. In 1991, The Israel Museum in Jerusalem pro-
duced a sumptuous catalogue for its exhibition of Urartian metal artifacts
(including a few replicas), consisting mostly of bronze, but including gold,
silver and iron objects, and some terracotta objects, edited by R. Merhav
the exhibitions organizer.3 Many objects exhibited and published reflect
almost the full range of Urartus metal artifacts, and their illustrations are
excellent. Succinct essays on metal deposits in Turkey and Armenia, and
inscribed metal objects were written by Belli; an essay on iron artifacts is
by R.B. Wartke; another on technology by A. Ruder and Merhav. Essays on
the categories of metal artifacts exhibited are by U. Seidl, P. Calmeyer, Mer-
hav, Kellner, R.-B. Wartke, K. Kohlmeyer, A. Ruder, and G. Zahlhaas. In some
essays excavated examples and their sites are mentioned.
All but a few (pp. 238, No. 31, 262, No. 10, 281, Nos. 68; and most prob-
ably pp. 240241, Nos. 35, 36, 259, Nos. 5, 6, and 307308, Nos. 35 from
Toprakkale,4 and p. 196, Nos. 52, 53 from Bastam) of the hundreds of objects
2 Kellner 1979, 152, Abb. A; Kellner 1980, 206, Taf. I, and Kellner 1993, pl. 61.2; also Born
and Seidl 1995, 54, Abb. 43.

3 Merhav 1991, Appendix II, 359361.
4 See Muscarella 1988, 28, note 4, 421, 429433.
624 chapter twenty
published are unexcavated. Presented is a corpus of largely plundered ob-

jects, most indeed from Urartian sitesbut plundered from where in mod-
ern geographical boundaries, Turkey, Iran, Armenia, is of course unknown.
(I have been able to recognize the birth site of only one of the many orphans,
see below.) Most objects in the exhibition are a loan of plundered material
provenanced in twelve museums in five countries, five are from antiquity
collectors and a number are from one or more dealers shops; Turkey is not
included among the lenders. That the one dealer identified was invited to
advertise his wares for sale in the Jerusalem Museum and in its Catalogue
suggests that one of the exhibitions purposes was to function as a bazaar
site offering antiquities for sale (not the first time this was done, see below)
at a museum site. Further underlining the core-problem addressed here, we
are unselfconsciously informed on a separate page that the exhibition was
made possible by a generous grant from an anonymous donor (I suggest
museum shorthand for a collector or dealer of antiquities), and by grants
from two prominent collectors, i.e. destroyers of this planets history, Jan
Mitchell and Jonathan Rosen.
Given the fundamental issue of the lack of site proveniences, one would
have expected the contributing scholars to address this matter. But most
accepted the required museum protocol: avoid facing up front this uncom-
fortable fact, except occasionally in a single sentence or phrase unrelated to
the essence of the discussion. Troubling is Calmeyers statement (on p. 23)
that many non-excavated helmets were probably pillaged (we do know
that they were not probably excavated), or Merhavs museum curators
euphemism (p. 202) referring to plundered objects as acquired through
trade. By whom?
For too long scholars have been conditioned, indeed much from museum
and collector publications and catalogues, to accept this museum-speak
code that keeps the details of the dirty business sanitized and discretely
set aside in the background; nice scholars are not expected to talk dirty
about precisely how the goodies were acquired through trade; plunder is
not a nice word. The Jerusalem catalogue and its scholar-contributors fol-
low this tradition. Throughout the essays one encounters other archetypal
dealer-curator clichs evidenced by the many statements that an unexca-
vated object was allegedly from X, said to be from X. Scholars here also
supply lists of various objects labeled found together in the region of Van,
reportedly found together, allegedly found together, or isolated orphaned
objects assigned to one particular tomb or one (archaeological) complex:
see pages 54, 71, 89, 100, 142143, 166, 178179, 185, 188, 223, 287, 294295, 298
299, figs. 7, 8, 13, 14.
Concerning the votive plaques (figs. 7, 8, pp. 287, 294295), Kellner easily
informs us that its provenience is known at least approximately to be near
Malazgirt (italics added), pace the captions for the two objects that give it as
Giyimli? a site not near Malazgirt. In an earlier publication, Kellner5 gen-
erously confided that his Malazgirt source was in Germany, the brother of
the man who ausgeackert the finds (whether at Gyimli or Malazgirt doesnt
matter), which included fibulae and bronze belts.6 The pedagogical message
here is that archaeology cannot function without the help of curators who
keep in touch with innocent peasants plowing their fields.
Kellner furthermore informs us (p. 287) that his Munich cache of objects
did not derive from a tomb, but probably from an Urartian settlement in
or near a fortification (found under a plowed field). These archaeological
provenience data are provided by a museum director while sitting in his
Munich office, and whose readiness and ease to share what he knows neatly
illustrates much of said to be Urartian metal artifact scholarship.
The Fundkomplex is a beloved term defined by those who use it as a
corpus of material painstakingly kept together along with the sites name,
by the innocent but archaeologically astute peasants who casually found
the objects, and then passed them to the honest dealers in Germany; but
smuggling and smugglers within and outside of Turkey are not mentioned.
In Merhavs catalogue there is a group of 10 Urartian silver and bronze
objects (pp. 206207, 217225), most inscribed with the names Ishpuini or
Menua (pp. 221225). They were earlier chronicled as deriving aus einem
Fundkomplex by Kellner who saw them in a Kunsthandelsgalerie in 1975,
and later described to be Ein sensationeller Fund 7 Writing in his office
in Munich, Kellner posited that they all probably derived from Inushpas
royal tomb, the locus of which he blithely identifies as probably Li, near
Patnos. Merhav, sitting in her Jerusalem office, disagrees;8 to her they derived
from a temple treasure not a tomb locus. Indeed, Li was plundered, then
excavated as a salvaged site, where a bronze belt was in fact recovered (see
below): but what has this to do with the material Kellner saw in Germany,
Merhav in Jerusalem? We are expected to be grateful to get such valuable
specific archaeological data generously provided by museum directors and
curators (not to mention dealers).
5 Kellner 1982, 84; Caner 1998, vii accepts the Malazgirt label.
6 See also Kellner 1991, No. 70.
7 Kellner 19751976, 57, 6667; Kellner 1991, 8; also Calmeyer 1986, 81, note 10; on this see
Muscarella 1988, 423, note 13.

8 Merhav 1991, 223; Barnett (Barnett 1982, 339, note 180) knows they came from Transcau-
casiaand indicate early Urartian penetration there.

626 chapter twenty
Fundkomplex? Found together? A more accurate archaeological term is:

Museumkomplex, objects plundered from unidentified sites we shall never
know about, selected and placed together to be sold by dealers, then pur-
chased by museum curators and directors, and serious collectors. That we
simply cannot know the cultural loci, whether from burials or occupational
sites, not to mention their geographical origin in Urartu, that is, whether
from North-Eastern Turkey, Armenia or North-Western Iran, of the objects
in the catalogue (except for a candelabrum, see below), is ignored or uncom-
prehended.9
Chance finds do occur, evidenced from time to time, for example the
salvage excavation from a chance find by workers at a car factory in Ereven,
Armenia, and brought to our attention by R. Biscione, where we learn of the
presence of bronze belts. Another, fairly secure such find, was confiscated
by authorities from local villagers who had obtained recently plundered
material from a construction work in the village of Burmagetiit, near Elazig;
here a number of bronze belts were recovered (see below).10
Further demonstrating how incorrect information is circulated by those
uncritically dealing with unexcavated objects is the fact that several objects
published in the Jerusalem catalogue, while ancient, are not Urartian arti-
facts: pp. 91, No. 50, 228, 240241, Nos. 37, 38. Concerning No. 50, a rampant
lion on a disc, Rehm11 could not cite a single Urartian parallel, but did not
(could not?) remove it from her 1997 museum catalogue. Merhav admits that
the bull head attachments, Nos. 37, 38, differ from proper Urartian attach-
ments and rejects an eastern Anatolian provenance; but this (correct)
statement does not explain why they were exhibited and published in her
Urartian catalogue (I suggest because she did not wish to offend their rich,
powerful Swiss owner). They were published as Urartian, from oost-Turkije
by Vanden Berghe and de Meyer.12 The helmet with a protome crest facing
forward (p. 128, No. 10) is Assyrian (as the caption sates), which attribution
T. Dezs and J. Curtis demonstrated;13 these authors (pp. 114115, pls. XVII,
9 For criticism of the common occurrence of Schatzfunde and the geschlossenen Fund-
komplex in Urartian scholarship, see Haerinck and Overlaet 1984, 54, 56; Muscarella 1988,
395396, 398, note 6, 423, notes 13, 14; Muscarella 2000, 146147, 213, note 54; Garrison 1994,
150; and Simpson 2005here with a neat discussion of museum anti-archaeological prac-
tices. Seidl (Seidl 2004, 17) correctly warns us about Komplexe, but then accepts others (Seidl
2004, 17, 65, note 37, 132).
10 Biscione 1994; Yildirim 1991.
11 Rehm 1997, 225.
12 Vanden Berghe and de Meyer 1983, 207208, Nos. 177, 178.
13 Dezs and Curtis 1991, 107119, figs. 1, 19, 20: excavated at Nimrud; see also below.
XVIIIfor the latter see below) also publish two other probably Assyrian
examples curated in German museums and in the catalogue labeled Urar-
tian by Kellner.
I further suggest that decoration on some objects exhibited are modern
forgeries, or at least to be considered suspicious (see also below): pp. 87,
Nos. 40, 41, 44; 92, Nos. 51, 52; 122, No. 8; 126127, fig. 8, No. 9; 129, No. 11 (a
helmet excavated at Ayanis, is of the very same form and bears an ancient
[sic] incised scene):14 166, No. 2; 167, No. 4. D. Collon15 suggests that the gilt
relief scene on the silver bucket, pp. 220221, No. 22, may be modernI am
not sure and reserve judgement; however, she does not indict the scene on
p. 129, No. 11. Problem pieces include pp. 106, No. 67, 108109, Nos. 7173.16
And there is naked dissimulation. In an essay on bronze candelabra, Mer-
hav mentions six excavated complete examples, four from Karmir Blur, one
each from Toprakkale and Altintepe (pp. 262267, figs. 1012, Nos. 10a, b). She
then introduces a bronze example curated in her institution, the Jerusalem
Museum (pp. 264, 270, No. 11), that has come to light (another museum
euphemism; here designating an artifact saved by simple peasants from
the dark earth! Like some of the examples mentioned, it is typically Urar-
tian), consisting of an upright shaft placed over a three-legged base, each
of which has a recumbent lion on the curved feet; here only two lions are
preserved, and the missing one has been restored (three candelabra from
Karmir Blur lack the lions). The candelabrum in Jerusalem bears an inscrip-
tion of Menua, son of Ishpuini (p. 357).17 Although the isolated lion excavated
at Kayalidere is correctly identified as deriving from a candelabrum (pp. 265,
274275, fig. 1), the equally excavated lion from Aznavur-Patnos, also dis-
associated from a candelabrum18 is not accorded an excavated attribution
in her text. In fact, it is only mentioned once (with no bibliographical ref-
erence), buried in a caption for one of the Jerusalem candelabrum lions
(p. 280, No. 5), where it is casually noted, Similar [sic] in shape and details is
a bronze lion found [not excavated!) at Patnos. The Jerusalem lion is shown
in front and back view, but not the side view, as accorded the Kayalidere lion.
There is more to consider. The candelabrum was donated to the Museum
in 1972, and in the previous nineteen years the museum had possessed it, it
14 Derin and ilingiroglu in ilingiroglu and Salvini 2001, 181, fig. 13.
15 Collon 1993, 126.
16 For earlier discussion and references of these objects, see Muscarella 2000, 148152; for
other forgeries of Urartian art see Muscarella 2000, 147154; Haerinck and Overlaet 1984, 53,
55 also raise the issue; see also below.
17 Also Salvini 1991.
18 Boysal 1961, 199, figs, 13.
628 chapter twenty
was to my knowledge never hitherto exhibited or published. The acquisition

date is totally omitted from the catalogues text; only mentioned is that it
was acquired in recent years (a revealing cover-up term). Bellis catalogue
essay (pp. 4748) refers to the Jerusalem candelabrum only in context of its
inscription; Salvini avoided any discussion of the lions relationship to other
candelabra or its acquisition history, noting only that it had been acquisito
anni fa.19
But one explanation exists for these conscious omissions, for the oblique
reference to the Aznavur-Patnos lion, the alleged listing of all the exca-
vated examples of candelabra from Urartian sites, the description of the
Jerusalem candelabrum, and keeping the acquisition date secret for almost
two decades: the lion missing on the candelabrum in Jerusalem is the iso-
lated lion excavated at Aznavur-Patnos, a fact long known in the Jerusalem
museum authorities, who ordered that time had to pass before it could be
published. And to admit this would have required a museum curator to
explain how her museums candelabrum came to light, to acknowledge
that it was plundered from Aznavur-Patnos, in Turkey. The Turkish authori-
ties and all archaeologists must be kept unaware of this museum secreta
truth that dare not be revealed in an essay said to be concerned about exca-
vated candelabra and archaeology.20 Seidl recognized the connection, refer-
ring to the Jerusalem candelabrums source as Aznavurtepe (?) and also
mit Wahrscheinlichkeit from that site; but she did not carry it further.21
Nevertheless, archaeologists excavating the catalogues multi-level text
can learn information about Urartian metal artifacts: that in fact eight
candelabra were recovered at Urartian sites. One was kidnapped and sold, at
which time its birth-name was changed, the Aznavur-Patnos candelabrum
was baptized the Jerusalem Museum candelabrum: a classic example of the
Museum Ritual executed to distort archaeological reality for a tendentious
cause. On the positive side, however, it has been demonstrated that scholars
can learn much about Urartian metals from a close reading and decoding
19 Salvini 1991, 344.

20 The Jerusalem Museum has no monopoly on Museum-Ritual dissimulation. I have
recorded the dissimulation (the kindest word I can think of) by a Metropolitan Museum
curator and Director. Jointlyin a museum publicationthey expediently transmogrified
Achaemenian objects into Greek artifacts, solely to cover up the true cultural provenience
of their purchase of the plundered so-called Lydian Treasurewhich, nota bene, when
exposed was eventually (and quietly) returned to Turkeynot to Greece: Muscarella 2000,
44. For specific case histories of museums deceit, lies, distortion and dissimulation of the
archaeological record to fit their agendas see ibid. 27, 2225, notes 1, 4, 5; and Simpson 2005.
21 Seidl 2004: 15, 25, C. 11, Taf. 3.
of museum (and also collectors) catalogues, far more than their editors
desired.
The data excavated from the catalogue at large concisely reveal the state
of affairs of Urartian archaeology, and specifically define the complex issues
that utterly compromise a study of the cultures metal artifacts. Not enough
scholars recognized the catalogues underlying message and archaeolog-
ical fictions. One example is the otherwise important historical work of
Salvini,22 one of the very best scholars of Urartian history and language, who
here demonstrates how a scholar can be so readily seduced by a catalogues
charm. In his discussion of Urartian metal artifacts (Die Bronzekunst chap-
ter), paradoxically only a few excavated artifacts are cited and discussed, and
plunder matters not at all.23 Privileged however, is the material published in
the Jerusalem exhibition catalogue, which to him is eine Summa unserer
Kenntnisse hinsichtlich der Bronzen und der urartischen Metallarbeieten
im besonderen.24 This claim is less a hyperbole than uncannily correct, but
only ironically, and not in its intended sense. One scholar who confronted
some of the issues raised here was M. Garrison in his review of the Jerusalem
catalogue. He recognized its usefulness and importance, but stressed the
non-excavated, plundered nature of the material exhibited, how this fact is
basically ignored, and that some of the contributing scholars run[s] the risk
of legitimating the very forces that have brought such havoc to the study of
Urartian art.25
The Jerusalem exhibition of Urartian artifacts together with its catalogue
was not the first such event; several other museum-generated exhibitions
anticipated it in form, content and spirit, albeit less extravagantly. German
museums in particular specialized in exhibiting Urartian artifacts accom-
panied with catalogues in the late 1970s and early 1980s, the first exhibi-
tion setting an example for the others. The first exhibition occurred in
Munich in 1976,26 where nearly 200 objects exhibited derived from muse-
ums, private collections, and dealers (the names of the latter two cate-
gories are identified only by initials), which set the pattern for all suc-
ceeding exhibitions. The accompanying catalogue also provides us with
many examples of angeblich or velleicht from X, or just plain absolute
Transkaukasien, museum-generated proveniences. Noteworthy was the
22 Salvini 1995.
23 Salvini 1995, 170182; on this issue see Krolls apt criticism, Kroll 1997, 204.
24 Salvini 1995, 171; see my review of Salvini, Muscarella 1997b, 737.
25 Garrison 1994, 151.
26 Kellner 1976.
630 chapter twenty
inclusion of excavated artifacts in the exhibition and catalogue (courtesy of

the then-in-place Iranian law sharing excavated material with the excavat-
ing institution, this one in Germany), consisting of 73 artifacts from Bastam
in North-Western Iran (only sixteen of which are minor bronze and iron
artifacts), superbly excavated by W. Kleiss; we are told that Turkish law pro-
hibited any material coming from their museums. The published exhibition
report is untypical for a catalogue, for there are very few illustrations of the
objects exhibited. Its long-term value remains the good scholarly articles
written by various authorities, the majority concerned with Urartian history
and archaeology in general, not specifically with the exhibited objects, many
of which are metal (only two articles deal with these). In the introduction
(p. 10) the museum Director H.-J. Kellner laments the plunder problem and
indeed, ingenuously coined a phrase to describe it aptly and accurately, eine
zweite Zerstrrung Urartus, seiner Kultur und Geschichte. But he is oblivi-
ous of the inconsistency unintentionally revealed here: that he himself (as
well as many other museum staff who have a passion to enhance their col-
lections), purchased the plundered artifacts in the exhibition, and that by
this act played a major role in the Zerstrrung, in legitimating, as Garrison
noted, the plunder and the subsequent scholarly havoc. That the passion to
purchase and to include dealers objects in the catalogue is the direct cause
of the plunder in the first place is evaded: but this is made manifest by the
Catalogues title, that the plundering of Urartu has resulted in the museum-
speak Wiederentdeckter Urartu.
Three more German exhibitions of Urartian metal artifacts were
mounted. One was in Hamburg 1979, another in Muenster in 1980.27 The
material exhibited was the same as in Munich, but without the Bastam arti-
facts. Both exhibition publications are the very same, with few metal or
other artifacts illustrated, and both presenting a good summary of Urar-
tian history and archaeology by S. Kroll, one of the foremost of Urartian
scholars. The other exhibition was in Berlin in 1984, consisting of Urartian
and Caucasian artifacts.28 Seventy-five Urartian objects were exhibited here,
all of which we are informed in captions and the object list, are ange-
blich aus Ostanatolien; and five of which aus einem zusammengehren-
den Grabkomplex stammen, and sollen aus dem [same] Grabkomplex
(pp. 3441, Abb. 1820, 2224, 30)but no references are given for the loci to
sport the museums phrases. A number of the objects exhibited here (along
27 Kroll 1979; Kroll 1980.

28 Kohlmeyer and Saherwala 1984.
with the non-archaeological posture) appear also in the Jerusalem cata-

logue. One is a bronze cauldron with two bulls head attachments of non-
Urartian style and form (and none like them has been excavated in Urartian
territory) set on an iron tripod stand of a form hitherto unknown. Note that
when the cauldron was exhibited in Ghent (below) no tripod was shown.
Because of an outline marking on the cauldron, the base was assumed to
belong to this particular cauldron, a juxtaposition I challenged in 1992.29
In 1983 an exhibition was produced in Ghent, Belgium.30 This exhibition
(some of which material appears in the Jerusalem catalogue) included many
metal artifacts from European museums, collectors and, again, dealers, but
also a number of excavated objects from Karmir Blur, Alishar, Toprakkale
(loaned by the Hermitage Museum), and a few from Bastam. For the unexca-
vated orphans, the authors inform us that most derived from Oost-Turkije,
others derived from noordwest-Iran, or Kirmanshah.
Still to be exhibited and fully published is a collection of almost 200
mostly Urartian bronzes first offered to the Metropolitan Museum in 1967 by
a Philadelphia dealer and later to the University of Pennsylvania Museum,
which purchased it the same year. The collection consists of bull caul-
dron attachments of non-Urartian form, also animal statuary, bracelets,
and apparent chariot units. E. zgen has published a few of the museums
objects in two articles dealing with chariot equipment.31 This purchase from
its provenance in Philadelphia is asserted by zgen to derive from Eastern
Anatolia. He lists them together32 with another group of objects deriving
from Munich, concerning which he informs us derived from a princely
Urartian tomb in Eastern Anatolia: his cited source is the field archaeol-
ogist H.-J. Kellner. Thus, we have two separate dealers hoards considered to
represent (only) two separate plunderings, both attributed to Eastern Ana-
tolia.
Another source for those interested in the probably pillaged back-
ground and long-distance travel of Urartian metal artifacts is a catalogue
29 Kohlmeyer and Saherwala 1984, 34, Abb. 18, 19; Muscarella 1992, 27; Seidl 2004, 132,
note 743 asserts that I did not read the Berlin text; she has. But ignored in her reading is
that both cauldron and stand are unexcavated, and that an outline for a stand per se would
be expected, but by itself (for no other information is given) does not demonstrate that the
Berlin stand and cauldron belonged together before it was plundered. Seidl did not challenge
the alleged Grabkomplex.
30 Vanden Berghe, de Meyer 1983.
31 zgen 1983, 117, fig. 18; zgen 1984, figs. 8, 24, 2733, 36, 45; see also zgens Ph.D. 1979
dissertation in the University Museum.

32 zgen 1984.
632 chapter twenty
published by The Ancient Orient Museum, Tokyo.33 Illustrated are 127 Urar-
tian metal objects purchased by Japanese private collectors, who are ac-
corded deep gratitude for allowing their possessions to be revealed (but
whose names are not vouchsafed). Indeed, the catalogue publishes a valu-
able bibliography of Urartian studies in many languages (985 entries), but a
number are of works that merely mention Urartu; for an extensive bibliog-
raphy see P. Zimansky.34
The distribution of Urartian artifacts excavated outside their homeland is
important for the study of cultural influence, political engagement or trade
beyond Urartus borders, particularly concerning the west, and metal arti-
facts have been the main source of research in most investigations of long-
distant exchanges. Since the 1960s I and H.-V. Herrmann have argued that
examples of large bronze cauldrons bearing bull or siren protome attach-
ments excavated outside the Urartian homeland, at sites in Turkey, Italy
and Greece, which many scholars had concluded over many years were
imports from Urartu (thereby demonstrating widespread knowledge of that
culture in Phrygia and the West) were in fact not Urartian productions. We
argued from a stylistic analysis that they were manufactured and exported
from elsewhere, probably from North Syria, specifically that the cited bull
attachments quite clearly are not the same in style and manufacturing
technique as the Urartian examples, and the sirens are not characteristic
of Urartian style.35 Merhav disagrees, and reprises the old interpretation
that these cauldrons recovered abroad indeed are of Urartian manufac-
ture, including some from Toprakkale, from whose workshops they were
exported West, to Gordion, Italy, and Greece.36 I have not changed my mind
on this important issueno cauldrons recovered at any site beyond Urartu
arrived from Urartu: of those published in the Jerusalem catalogue, some
are Urartian, one bull attachment (p. 238, No. 31) was excavated, two oth-
ers, unexcavated, most probably have an Urartian background (pp. 240241,
Nos. 35, 36). Of the other plundered examples, Nos. 30 and 32 (pp. 236, 238),
are Urartian (Nos. 37 and 38, mentioned above, are ancient, but not Urar-
tian). And to me, the cauldron with achments in Munich, is not Urartian
33 Tanabe et alii. 1982. For a critique see Muscarella 1992, 3132.

34 Zimansky 1998.
35 See Muscarella 1992, 1635 for a bibliography and summary, also Salvini 1995, 176178;
and Seidl 1988, 169 and note 1.

36 Merhav 1991, 229235. As for siren attachments alleged to be from Toprakkale: in
Muscarella 1988, 28, note 4 and Muscarella 1992, 22. I made it clear none derived from that
site: pace Seidl 2004, 3, note 7, where she claims the contrary; she only cited a much earlier
(1962) article of mine.
(pp. 242243, No. 39); it was plundered and has no known archaeologi-
cal provenience. The two terracotta cauldrons with bull head attachments
(pp. 239240, Nos. 33, 34) also have an Urartian background.
In the context of recognizing and defending contacts between Urartu and
the West as active historical events, the museum curator R.D. Barnett leaps
(very high) when he suggested that the throne of King Midas of Phrygia
sent to Delphi could have been made, not by Phrygians, but by Urartians;
that King Urpalla at Ivriz is decorated with Urartian design encrustations,
which designs are also to be seen on plaques excavated in the Ephesian
Artemisium: all of which is serious, indeed embarrassing, distortions of the
archaeological/historical record.37
Up to the present only one, or possibly two, examples of Urartian artifacts
have been recovered outside of the homelandif, of course, my attribution
arguments about the cauldron attributions are not contradicted by future
excavations. One, manifestly, is a bronze bell from a sanctuary on the Aegean
island of Samos (B474), which has a number of inscribed Urartian examples
as parallels.38 The other is a statue also excavated at Samos (B1217), which
a number of scholars argue is Urartian; I am not certain because its style
is not easy to determine, but cannot categorically refute the attribution.
The Samos bell, and apparently the statue derived, ultimately at least, from
Urartu; how they got from there to a sanctuary on Samos is as intriguing as
it is unknown.
A number of publications report on the occurrence of unexcavated Urar-
tian metal artifacts from loci outside of the states known boundaries, or
identify their derivation from sites within the Urartian sphere. I refer here
to scholars who know with ease the precise site or geographical location
of plunder-derived, smuggled artifacts shown to them by dealers or collec-
tors, and share this information. Such determinations most often are gener-
ated in museum offices. It is not necessary to cite every occurrence of this
Leitgedanke, common, practice; a few examples will suffice to document the
wide parameters available to bazaar archaeologists.
In two publications, R. Ghirshman39 shared with colleagues what to him
was an important archaeological discovery. He reports that five Urartian
objects, blinkers and horse bits, all but one inscribed, housed in an Iranian
37 Barnett 1982, 369, 371.

38 Muscarella 1978, 6365, 66; Muscarella 1988, 427428. Years ago on a visit to Samos I
asked archaeologists there to clean the bell to determine if it might reveal an inscription; I
have had no further news.
39 Ghirshman 1964, Ghirshman 1966.
634 chapter twenty
private collection (Foroughi) and in a Teheran dealers shop, were trou-

vailles, faites fortuitement [sic] par les paysans. These, to him, casual and
chance finds (plunder is a dirty word never spoken by Ghirshman) provide
precious data to be used to obtain knowledge of ancient Urartian expan-
sion, which pedestrian archaeologists have been unable to discover, For,
inasmuch as Ghirshman knows (and therefore doesnt have to explain fur-
ther) that they derived from tombs located southwest of the Caspian Sea,
he is obligated to share with colleagues this historical and chronological
information about Urartian contacts with this region, hitherto not known
to have been within the Urartian orbit. One of the uninscribed objects, how-
ever, a corroded paddle-shaped object, is neither Urartian, nor a blinker,
although accepted as such by Merhav.40 Ghirshmans fantasy Caspian Sea
locus was unhesitatingly accepted as if it were an archaeological report by
several Urartian scholars: Azarpay, Zimansky, Salvini, Merhav, and zgen.41
G. Gropp, an epigrapher, correctly challenges Ghirshmans fabricated
provenience to the Amlash region.42 But he then continues with the same
methodology, ascribing a group of twenty bronze objects, equestrian equip-
ment, some inscribed, curated in a German museum and a private collec-
tion, to one grossen Fundkomplex, plundered (his word) from a deposit at a
locus near the Turkish-Iranian border.43 Gropp seems at first to have doubts,
but then appropriately cites Kellner, the authority on locus and Fundkom-
plex matters. End of discussion; we have here for archaeological study a
Fundkomplex. Because some of the objects have traces of burning (pp. 98
99), Gropp suggests that the group he has assembled in Germanyremains
of a chariotderived from a temple depot or palace rather than from a
grave: we now have all the data we need to write an article on an Urartian
chariot recovered from an intuited specific cultural locus.
Curator W. Nagel of the Museum fr Vor- und Frhgeschichte, Berlin,
reports that an inscribed Urartian helmet in his museum bears a museum
labelwhich piece of paper must be obeyed for it is curated in his museum
and therefore is an archaeological recordasserting it derived from
Rutchi Tig, a site in the Caucasus.44 Moreover, another curator-generated
40 Ghirshman 1964, 60, fig. 9; Merhav 1981, 6465, No. 38.

41 Azarpay 1968, 12, Zimansky 1998, Nos. 961, 1006 (elsewhere in this publication are many
said to come from loci references: see Muscarella 1988, 421, 423424, note 14; Muscarella
1997a: 207), Salvini 1980, 186, notes 32, 33; Merhav 1981, 65; and zgen 1984, 99100, 101.
42 Gropp 1981, 95, note 1.
43 Gropp 1981, 97, 99; Haerinck and Overlaet 1984, 5660 confront this solecism accurately.
44 Nagel 1959, 60. Nagel was not reluctant to reveal his skills in recognizing the loci of
other objects curated in his museum. He knew immediately when he saw and purchased
archaeological fact is that the helmet belongs to a Fundkomplex associ-

ated by him with other objects housed in other museums. Kellner, and
Kohlmeyer and Saherwala accept the Caucasian attribution; L. Vanden
Berghe and L. de Meyer at least placed a? after this attribution; Zimansky
correctly records the helmet as from an unknown provenience.45
Another museum curator, H. Hoffmann, employs the same museum-
speak formula in his publication of an Urartian belt purchased from a dealer
and curated in his Hamburg museum: It is said to have been found near
the modern town of Diarbakir. We are not informed who said it, but it
was surely the vendor, whose archaeological authority compels Hoffmann
to accept the provenience, as does Kellner.46
The Turkish museum curator O.A. Tasyrek published a number of
bronzes known from irregular excavations and acquired by him (appar-
ently by purchase) at his museum in Adana (which is not in the vicinity of
Urartu). Based on information revealed to him by antiquity dealers in Van
and Adana, he assigns them to specific Urartian areas or sites, from near
Patnos and at Aznavurtepe in the same area; and several belts from
the Dedeli village of Patnos county; for one example, he records that It
has been established with no further explanation, that the belt was found
in the Dedeli village.47 The latter attributions were readily embraced by
Kellner,48 who further informs us that they derived from a Grabfund (a
belt was excavated at Dedeli, see below). Elsewhere Tasyrek also reports
that an unexcavated bronze belt derived from Karahasan, south of Malaz-
girt (a site we have encountered above), where we are told it was found
(Nagel 1963, Nagel 1970) in German dealers shops that two statuettes derived from one
Fundkomplex, Hacilar. After they were exposed by me as obvious modern forgeries, his
archaeologist wife defended them vigorously as indeed deriving from ancient Hacilar: both
knewand therefore did not think it necessary to provide us with comparanda: perhaps
because none exist. Nor did they deign to discuss in print the manner in which the said-to-
be plundered neo-Hacilar figurines left Hacilar, Turkey, and reached Berlin: for details and
bibliography see Muscarella 2000, 136137, 438439, Nos. 13, 14. The present Berlin Museum
curators and director continue to obey the museum code, for at the time of this writing
(2005) both figurines are still on exhibitto protect an honored curator, the museum, and
to educate visitors about neo-Hacilar, which in Berlin is merely a semantic difference from
ancient Hacilar.
45 Kellner 1980, 206; Kohlmeyer and Saherwala 1983, 38, Abb. 26; Vanden Berghe and de
Meyer 1983, No. 16; Zimansky 1998, No. 946.

46 Hoffmann 1971, 701; Kellner 1977, 434; Kellner 1991, 5, 21, 80but I could not find it
catalogued.
47 Tasyrek 1975a, 151, note 3; Tasyrek 1975b, Nos. 1, 19; Tasyrek 1977, 157.
48 Kellner 1991, Nos. 63, 102, 103; the Dedeli attribution for these were rightly challenged
by Curtis (Curtis 1996, 121). Note that a bronze belt was excavated at Dedeli, see Kellner 1991,
No. 373 (below).
636 chapter twenty
by local peasants. It is now known as the Karahasan belt.49 In yet another

article Tasyrek50 reports on twenty-three silver and bronze vessels, also
acquired by his museum at Adana. Inasmuch as a dealer told him that the
silver vessels were found together with some inscribed bronzes bearing the
name Sarduri son of Argishti, he assumes they all must have come from that
rulers tomb. Who translated the inscriptions, and where are these inscribed
bronzes now? Merhav casually repeats Tasyreks Sarduri tomb provenience
revelation; but it was correctly challenged by Garrison.51
How does one explain why a purchased Urartian bell provenanced in the
Ashmolean museum and published there as from Western Iran was altered
by zgen52 to North-Western Iran, an equally made-up provenience? This
criticism is not trite, for it shows yet again how anyone can casually assume
the right to create and to recreate the loci of unexcavated objects across the
map. Another, and egregious, example of this bazaar archaeology method-
ology is an article written by A. Piliposyan.53 Herein we are informed for
the first time ever of a major and important find of Urartian artifacts at
sites in Syria, specifically belts at Tell Brak, and another at Tepe Leilan,
finds missed by archaeologists excavating these two sites. Piliposyan (and
the journals editors who also bear responsibility here) expect us to accept
in good faith this fantasy report as an archaeological scoop. The Tell Brak
belts are not discussed, with no explanation, but the other had been found
or, reportedly found at Leilan. Piliposyan doesnt bother to inform his
colleagues where he saw the belts, but his Leilan informer was surely the
unnamed private collector (in Armenia?) cited as its present owner. To
interpret his sensational discovery and explain how bazaar-excavated belts
reached Syria in the 8th century bc (his date), the author provides an his-
torical reconstruction: Urartian armies penetrated into Syria and fought
battles there, events hitherto historically and archaeologically unrecorded.54
A few more examples of the methodology that produces archaeologi-
cal conclusions based on unexcavated metal artifacts will suffice. Again
they demonstrate how scholars sitting in museum offices automatically
and unhesitatingly utilize dealers inventories viewed in Europe to gener-
ate what they believeand expect others to believeare normal, bona
49 Tasyrek 1973, 204, Kellner 1991, No. 37 (another belt is listed as umgebung von Malaz-
girt, No. 70, mentioned above), Collon 1993, 127.

50 Tasyrek 1976.
51 Merhav 1991, 223; Garrison 1994, 150.
52 zgen 1984, 109, No. 1, note 136.
53 Piliposyan 1996.
54 See also Muscarella 2000, 147.
fide archaeological data. The German museum curator M. Maass (Badisches

Landesmuseum, Karlsruhe) in a lengthy article presents material on loan,
and exhibited in his museums vitrines (some of which he later purchased),
eighteen pieces bisher einzigartige Kunstwerke which are a zusammen-
hngender Fund55 One is an Assyrian-form helmet (originally published
by Kellner as Urartian but correctly attributed by Maass), with an animal
head protome, a Raupenhelm, that has an Assyrian-like tribute-bearing
scene inscribed on the base.56 I believe that the Assyrian scene was recently
added, is a modern forgery.57 Kellner, and later T. Dezs and J. Curtis,58 how-
ever, accept the scene as an ancient-made Assyrian representation. Because
Maass knows (the dealer told him!) that all the objects he is curating derived
from an Urartian Komplex, he is compelled to instruct colleagues and stu-
dents how an Assyrian helmet with an incised royal scene (which, not inci-
dentally, does not occur on any other similar helmet) came to be deposited
with Urartian objects in an (Karlsruhe) intact hoard. Given that he knows
that the dealer (or his source) didnt just select it from his inventory of strays
plundered from different sites in the Near East and casually place it with the
hoard offered for sale, it ineluctably remains to be interpreted as a unit of the
hoard: the helmet must have been either gifted in antiquity to the Urartian
court by Assyrians, or it arrived in Urartu as booty from a battle. Therefore
one is obliged to catalogue the helmet along with Urartian metal artifacts:
Urartians touched it. Even if one thinks the scene is ancient, this interpreta-
tion along with the Komplex claim is unacceptable. I also have doubts about
a scene on another helmet Kellner publishes, a male mastering two floating
winged bulls.59
His colleague E. Rehm60 also published in a Karlsruhe museum catalogue
the same helmet (in her Mesopotamian section), here furnishing a draw-
ing of the scene (but which to my eyes is not an accurate reproduction of
the actual crude execution involved). She also repeats Maass claim that
other objects (but she adds a soll) were found together with this helmet.61
55 Maass 1987, 65 and following.

56 Kellner 1980, 211212, Taf, XIIIXIV; Maass 1987, 6571, Taf. 2, 3.
57 Muscarella 2000, 150, No. 19. I suggest that another Assyrian scene that has been
recently incised to a helmet is Born and Seidl 1995, Abb. 122, 96, 97: ein Unikum, (p. 43);
see also Muscarella 2000, 184, No. 42.
58 Kellner 1980, 211212; Kellner 1993, 326, pl. 6063; Dezs and Curtis 1991, 115116, pl.
XVIII.
59 Kellner 1980, 210211, Taf. XIXII; accepted as ancient by Born and Seidl (Born and Seidl
1995, 96, Abb. 85); see Muscarella 2000, 152153, No. 31.
60 Rehm 1997, 105108, note 36, fig. 1, Abb. 196199.
61 Rehm 1997, 105, note 36.
638 chapter twenty
Following her colleague concerning how to recognize the manufacturing

locus of a helmet purchased in Karlsruhe, she repeats his interpretations
regarding gifting or booty. Proceeding further, she presents another exam-
ple of museum scholarship ritual: publish and positively interpret every
object your museum has purchased. Singling out, correctly (but not suffi-
ciently) a number of flschlicherweise executed representations, she raises
the possibility (Dies ist durchaus mglich) that these technical errors may
be explained by perceiving that the scene could have been created in Urartu.
Rehm calls our attention (in footnote 46) to a Nimrud ivory decorated with
einer hnlichen Szene. That this genuine scene might in fact be an argu-
ment against the alleged ancient execution on the Karlsruhe Museums
scene, and that the mistakes noted were probably (or at least could have
been) made by a modern provincial craftsman, does notcannotcome
into the discussion: because having been purchased by her museum, the
scene is patently genuine (see also Calmeyer, below). Museum-generated
interpretations of its curated goodies have patently uncovered an example
(albeit rare) of an (apprentice) Urartian craftsman reproducing, as accu-
rately as was possible for him, an Assyrian scene. Intuitive analysis thus
allows the helmet to be introduced in a museum catalogue as an Urartian
metal artifact recovered recently from an Urartian site. Furthermore, from
this unexpected (because hitherto not revealed by archaeologists digging
slowly in Urartu) discovery we now knowfor the first time everthat
Urartian craftsmen could be and indeed were directed to decorate a foreign
helmet with an appropriate foreign-inspired scene, and in the manner of
that foreign style: all new cultural information about Urartian metal work-
ing strategies is revealed from an object excavated in a German bazaar.
Maass presented other pieces that to him were found together with the
others cited above at an ancient site, that are probably modern forgeries.
Particularly disturbing is the decoration on a horse blinker and on a breast-
plate (p. 88, no. 18, Taf. 8).62 Equally, the incised decoration on a helmet from
a Swiss collection that Maass also cites as being a unit of his discovered Kom-
plex is modern.63
In 1980 Kellner published a series of helmets as deriving from the same
Sammelfund, one of several such collections seen by him in the Karls-
ruhe bazaars, which to him are readily accessible sources of archaeological
62 Maass 1987, 84, No. 13, Taf. 9,2; 88, No. 18, Taf. 8; Muscarella 2000, 148149, Nos. 15, 16,
and 150151, Nos. 2024.

63 Maass 1987, 69, note 23. It is the same helmet as Calmeyer in Merhav 1991, 129, No. 11,
mentioned above, and Muscarella 2000, 150, No. 18.

knowledge. One helmet is decorated with a series of decorative scenes, some

of which I suggest are modern.64 Calmeyer published this helmet again,
wherein he correctly listed all the many aberrations in the decoration
but which aberrations to Calmeyer were merely wohl Fehler des [ancient]
Handwerkers aus Ost-Anatolien, who were Kollegen of the ancient
Karmir Blur artisans.65 An alternative interpretation by a pedantic archaeol-
ogist would state that perhaps a modern craftsman accomplished the errors
and failures correctly noted. Another helmet published by Kellner has a dec-
orative scene that is also completely modern, mentioned above (see Maass
and Rehm).66
Kellner also published here an Assyrian helmet from the same source.
That this unexcavated helmet, decorated with a standing deity he believes is
Ishtar, derived from eastern Anatoliafrom Urartu, he immediately knew,
and he purchased it on the spot: regarding the locus, kann ich eigentlich
keinen Zweifel haben.67 He further reports to us from his Munich excavation
project that the Assyrian helmet, found in Urartu (not bought in Munich),
must have been taken home by an Urartian soldier as booty following a
battle with Assyrians68 (see also Rehm above).
All these publications were written to inform colleagues, archaeologists
and students how easy it is for attentive curators to reconstruct Urartian and
Assyrian history from inventories in European dealers shops.
The message is clear: why bother with excavation reports, with its pedan-
tic record keeping, where one finds so little material (local poor peasants
do a better job). Bazaar archaeological methodology as practiced by dealers
and curator collaborators is better capable of supplying so much impor-
tant data and cultural informationand far quicker than accomplished by
archaeologists. Alas, the message has been accepted by many scholars.
As with artifacts published in the Jerusalem catalogue, it is not uncom-
mon to encounter orphaned non-Urartian metal artifacts excavated in deal-
ers shops that are casually attributed by scholars to an Urartian workshop
and integrated by them into that cultures artifact metal repertory. For exam-
ple, a bronze mirror purchased by the British Museum curator Barnett was
baptized by him in London as Urartian; it has a rim decorated with incised
64 Kellner 1980, 207208, Taf. IIVI, 210; it is the same as Calmeyer in Merhav 1991, 126127,
No. 9; published as genuine also by Born and Seidl 1995, 2627. See Muscarella 2000, 149, 460,
No. 17; Seidl 2004, 7072, Abb. 32.
65 Calmeyer 1986, 81, Abb. 4, Taf. 1519; see reference to Muscarella in note 56.
66 Kellner 1980, 211212, Taf. XIIIXIV; Muscarella 2000, 150, No. 19.
67 Kellner 1993, 325.
68 Kellner 1993, 330.
640 chapter twenty
striding animals and addorsed calves heads at the handle juncture. Of the
animals, he knows they are Urartian although he casually notes, there is
little at present to compare them.69 He shares with us, however, that the
addorsed heads find a close parallel to a stone example in the Louvre
said to have been acquired in 1898 from the region of Van (he doesnt tell
us who said it). A bronze ladle, excavated at Tell Farah, Israel, was published
as an import from Urartu by the archaeologist R. Amiran solely because of
an apparent stylistic accord of its addorsed heads with the British Museums
alleged Urartian heads.70 Amiran then utilized the Urartian ladle exca-
vated in Israel as evidence to suggest that it might prove that the siren
cauldron attachments excavated at Gordion, Phrygia, indeed derived from
Urartu. Here an archaeologist accepts instructions to see around corners, to
recognize how unexcavated, but perceptively recognized, Urartian metal
artifacts are capable of revealing ancient activities, here otherwise elusive
long distance trade contacts.
The ladle and mirror were accepted as Urartian by Zimansky, who fur-
ther calls to our attention that Barnetts mirror is the first Urartian mirror
known.71 He also cites T. Kendalls publication of another unexcavated, pur-
chased Urartian bronze mirror, in the collection of the Museum of Fine
Arts, Boston;72 presumably, we should accept this as the second Urartian
mirror known. Kendall immediately recognized it to be Urartian, but did
not consider it relevant to note that there are no Urartian parallels. Despite
these museum attributions, both objects are not Urartian, neither in style
nor manufacture, and no mirror or similar ladle has been excavated to date
at an Urartian site. Also, a bronze belt published as Urartian by Merhav,
which Urartian attribution is accepted by Zimansky, was surely not made
in Urartu, but probably somewhere in the Caucasus.73 Tasyrek74 also eas-
ily recognized several objects in his Museum, all acquired from nomadic
dealers, as being Urartian artifacts: a bronze lion, stag figurine, and a female
statuettefound, according to a dealer, in the province of Mus, and naming
the exact locus. Tasyrek also identified a terracotta bull figurine as Urartian;
he presented no Urartian parallelsbecause there are none.
The Urartian metal industry produced a prodigious quantity of metal
artifacts (minus, of course, those discussed just above), as much as any
69 Barnett 1965, 52.

70 Amiram 1966, 90.
71 Zimansky 1998, Nos. 982, 1003.
72 Zimansky 1998, No. 891; Kendall 1977, 46, 48, fig. 15.
73 Merhav 1981, 6668, No. 39; Zimansky 1998, No. 767.
74 Tasyrek 1976, 105.
ancient culture known to us. Most forms are conveniently presented and
illustrated with bibliographies in the essays dealing with specific categories
in the Jerusalem catalogue, and also in the other catalogues discussed above.
For a chronological study, one should begin with M. van Loon, who not
only for the most part uses excavated material, but also intelligently covers
the range of metal artifacts:75 also G. Azarpay, who concentrates on some
excavated material, but doesnt discuss the full corpus.76 Then one goes to
Seidl 2004. Also, of course, the examples of excavated metal artifacts from
specific sites, viz., Ayanis and Karmir Blur.77 I record here in list form the
metal artifact corpus available to us: utensils made of bronze, iron, silver and
gold: vessels of many forms, bowls, jugs, askoi, beakers, situlae, cauldrons
with bull attachments; small bronze statuary in the round (rare), sculpture
associated with thrones, furniture and candelabra; jewelry of bronze, iron,
silver and gold, fibulae, straight pins with figured heads, body ornaments
such as earrings, brooches, bracelets and armlets, medallions, pectorals;
weapons of many kinds made of bronze and iron: numerous swords, daggers,
spears and arrows, helmets, shields, quivers, breastplates; many bronze belts
most decorated with iconographic or geometric scenes; votive plaques of
different shapes, but best known from the many hundreds associated with
Giyimli; horse and chariot gear of great forms and variety: yoke and wheel
equipment, studs, horse bells, breastplates, blinkers, nose plates, bits, etc. A
good number of these have excavated backgrounds.78
To succinctly exemplify the dilemma facing anyone studying these arti-
facts even within the excavated corpus, I cite a few prominent examples that
illustrate the confusion and difficulties involved when one seeks informa-
tion about proveniences and cultural contexts.
Many hundreds (Caner 1998 catalogues 813 examples) of unexcavated
fragmented bronze plaques, some decorated in Urartian style and iconog-
raphy, but others not, have been associated by many scholars with the
plundered site of Giyimli, south of Lake Van.79 Giyimli was subsequently
75 van Loon 1966, 80130.

76 Azarpay 1968, for which see Kroll 1997, 204.
77 Derin and ilingiroglu in ilingiroglu and Salvini 2001; Piotrovsky 1970, Nos. 3887. For
a survey of sites with metal artifacts see Seidl 2004, 615.

78 Tracking bibliography for the corpus is formidable, given the many venues. A bibliog-
raphy of Urartian metal artifact publications and sites where artifacts have been excavated
is in Seidl (1988, 1, partial, and 2004). Specific essays concerned with metal artifacts at large
include Merhav 1991, Zimansky 1998 (199227), and Derin and ilingiroglu in ilingiroglu &
Salvini 2001. A dazzling bibliography of Urartian metal artifacts exists in Rehm 1997: 165323,
but is very difficult to access as it is embedded in countless (repetitious) notes.
79 Kellner 1982, 80. Caner 1998 is a lengthy, painstaking description of hundreds of plaques
642 chapter twenty
excavated by A. Erzen, who recovered only a few fragments that conform

to the plundered material.80 I previously discussed the problems associated
with this material,81 how many known pieces actually derived from Giyimli,
and how many seem to be ancient. As for the first issue, we have Kellners
proclamation, based on private sources in Germany (all dealers), that other
(plundered) sites in eastern Turkey also yielded Giyimli-like plaques.82 Rehm
states it slightly differently and more viably, noting that we cannot assume
that all (genuine) Giyimli plaques were made there in antiquity.83 How
then can one judge which of the many strange, crude, and unparalleled
unexcavated Giyimli pieces are genuine, and which are modern forgeries,
inasmuch as only about two of the few excavated bronzes relate to the many
unexcavated plaques?
Perhaps the most ubiquitous of all Urartian metal artifacts, excavated or
not, are the bronze belts. Their ancient and modern history are known to
us from the publications and bibliographies of Azarpay, Tasyrek, Kendall,
and Kellner, Curtis, Rehm, and recently, Seidl.84 In Kellners 1991 extensive
catalogue, 449 excavated and unexcavated belts are listed, most illustrated.
It is a valuable and painstaking contribution, collecting so many examples
in one venue for reference; other excavated and unexcavated examples are
to be added, viz. Erevan, Burmageit, see below, and those in the British
Museum.85 Collons perceptive review of this catalogue states precisely what
is wrong, what limits the value of the work. She isolates and challenges Kell-
ners methodology, his unsupported conclusions regarding chronological
determinations, votive function, and gendering of the belts: because they are
manifestly based on a majority of unexcavated examples, for which Kellner
gives prominence to unprovenanced and looted material.86 In other words,
Kellner privileges the unprovenienced examples, and throughout his book
catalogued by motifs: it is filled with detailed descriptions and far-ranging comparanda; it is

not an easy read. Rehm 1997, 167201, 235323, Abb. 417690, presents a detailed discussion
of the background and description of the plaques and publishes a large corpus housed in
Karlsruhe. Now add Seidl 2004, 169197 (note that examples in the Jerusalem Lands of the
Bible Museum are incorrectly listed by Seidl 2004, 170, as Slg. E. Borowski).
80 Erzen 1974.
81 Muscarella 1988, 424, note 15, and Muscarella 2000, 155156, 461.
82 Kellner 1982; Caner 1998, vii. A plaque was excavated in Armenia at Armavir, Kellner
1982, 8283, Abb. 2.

83 Rehm 1997, 170.
84 Azarpay 1968, 4751; Tasyrek 1975b, Kendall 1977, Kellner in Merhav 1991, 142161, and
especially his catalogue (Kellner 1991); Curtis 1996, Rehm 1997, 203209, and Seidl 2004, 133
168.
85 Curtis 1996.
86 Collon 1993, 127.
is blissfully unaware of the distinction between what is known and what

dealers tell him; his conclusions are founded on an absence of empirical
archaeological foundation. In short, nothing over the years has changed his
thinking.
I have spent much time attempting to record how many Urartian sites
yielded belts and how many derived from each site. I did my own check.87
I have been able to document that approximately 45 examples have been
excavated from at least 25 sites; some of these finds occurred after Kell-
ner wrote his 1991 catalogue. He catalogued most of the then-known sites
where belts were excavated (p. 87; his book helped my research), but unfor-
tunately he did not list them in a separate category, included made-up site
attributions, and he did not give the number of belts recovered from each,
an important omission.
I list here the sites yielding belts and the number from each (given in
parentheses). From Turkey: Toprakkale (2, possibly 4), Altintepe (2), Adil-
cevaz (3; Kellner mentions two), Kayalidere (1), Dedeli (1), Cavustepe (1 or
more; not listed in Kellner 1991), Giyimli (2 fragments; we ignore the Giy-
imli? attributions assigned to many unexcavated belt fragments published
by Kellner in 1991, p. 87which provenience is correctly challenged by Cur-
tis 1966, 121), Burmageit (7 or more), Igdir (1), Li (1). From Iran: Guschi
(1).88 From Armenia: Karmir Blur (6), Erevan (3), Nor-Aresh/Arinberd (2),
Leninakan-Shiraq (1), Ani Pemza (1), Chrtanoc (1), Zakim (1), Muchannat-
Tapa (1), Tigranaberd/Zod (1) Mecamor (1). From Osseti: Tli (3). Belli89 men-
tions, in addition to some recorded here, belts from Maliska: which I have
not seen, but they should be added to the site list (for Umgebung von Malaz-
girt: see above). Kendall90 claims that belts have been recovered more fre-
quently in funerary contexts, an assumption that seems to be accurate;
the one from Cavustepe was not recovered in a mass burial as he states
(p. 31); and his fig. 3, p. 33, is not a belt.
Another common Urartian metal artifact is the straight pin with figured
heads in the form of animals, plant forms and geometric motifs. Zahlhaas91
briefly discussed excavated examples in the Jerusalem catalogue, together
87 See also Curtis 1996, 120121.

88 In Muscarella 1988, 433535, I published a belt fragment in the Metropolitan Museum
that is one of a group of similar examples, one of which was the Guschi belt, whose chronol-
ogy I discuss. I thought I made clear the former fragment belonged to another belt, but once,
inadvertently, referred to it as the Guschi piece. Seidl 2004, 7 caught the error.
89 Belli 1997, 295.
90 Kendall 1977, 30.
91 Zahlhaas in Merhav 1991, 184194.
644 chapter twenty
with some unexcavated examples: some of which were reportedly [by

whom?] found together in the same burial context (Nos. 4144). Merhav has
provided a good study and bibliography of the pins.92 She lists 18 excavated
examples from seven siteseighteen of a total of 260 pins known to her; 26
(five from excavations at Adilcevaz and Karmir Blur) are made from precious
metals, the others of bronze. How many other pins are tucked in collector
and museum vitrines we cannot say. An important addition to the corpus of
excavated examples comes from the excavations of burials at Karagndz,
which also included bracelets and weapons (these, together with the pottery
recovered, indicate clearly that the burials are typically Urartian in date, not
earlier, so-called early Iron Age).93 But the relative number of excavated to
plundered pins is meager, and typical.
One trusts that the above discussion has demonstrated that the first step
for objective Urartian metal research cannot begin with museum exhibition
catalogues and other publications critiqued above: these are secondary lit-
erature, but which indeed, must eventually be read and analyzed critically.
The subject can only be approached first through the primary literature,
that of the excavated sites yielding Urartian metal artifacts. The formidable
problem encountered, however, is that there are many excavated sites in
several modern countries, viz. Turkey, Armenia, Iran, where publications
are either scattered in many venues, or the sites remain inadequately pub-
lished or unpublished across many years. In addition, aside from modern
excavated sites, ancient Musasir justly merits consideration as an excavated
site, inasmuch as Sargon II of Assyria published a neat site report. But here
the problem for scholars is how to relate the anciently recorded artifacts
deriving from an Urartian temple to a study of Urartian metal artifacts.
The sole evidence for Urartian artifacts (most probably existing in northern
Mesopotamia) is Sargons relief depicting the temple captured during his
well-known VIIIth campaign in 714bc.94 Revealed here is valuable archae-
ological information, in particular, the depiction of shields with lion pro-
tomes attached to the temples walls. There is also the well-known contem-
porary tablet (archaeological) report of Sargon listing the mass of plunder
he acquired at Musasir: for example, six gold shields with lion protomes
from the temples cella. This shield form and its temple context (also as
noted, in situ on the walls there) are wonderfully paralleled (by excava-
tion!) in bronze, at another Urartian site, Ayanis, on Lake Van to the north
92 Merhav 1995.
93 Sevin and Kavakli 1996, figs. 2627.
94 van Loon 1966, 4344, figs. 5a, b; Salvini in Merhav 1991, 11.
(see below). Sargons catalogue of artifacts, his booty, lists enormous quan-
tities of metal artifacts,95 among them hundreds of bronze shields, vessels,
chariots, weapons, and so forth; also a silver bowl of Rusa is mentioned
(see also the silver bowls in the Jerusalem catalogue inscribed with royal
names: pp. 224225, 356, Nos. 2734). Kendall and Barnett believe that all the
objects inventoried by Sargon were Urartian productions.96 Salvini claims,
probably correctly, that Sargons catalogue is a precise listing of the booty
taken; and further, that while objects from different cultures are specifically
reported, most of the temples contents were Urartian productions.97 Belli
more cautiously posited that one cannot draw far-reaching conclusions
from the catalogue [regarding] the material culture of Urartu, as many
of the objects listed derived from other culturesTabal, Assyria, Habhu,
attributions mentioned in Sargons text.98 Bellis conclusion is sensible and
viable, it cannot be dismissed, for we know that members of several polities
dedicated votives to the Urartian deityor that booty from these lands was
dedicated by Urartians there. Rusas silver bowl and statue with chariot, and
the lion protome shield are obviously Urartian metal artifacts, manifestly
documenting Urartian productions here. But it cannot be determined how
many of the other recorded artifacts may be accorded the same attribution.
Some important recently excavated metal objects merit consideration.
From Ayanis (yes; not said-to be-from), a site excavated for years by A.
ilingiroglu, derives precious archaeological information and data on Urar-
tian architecture and metal artifacts, carefully excavated and their specific
loci recorded. A good number of bronze and iron artifacts have been recov-
ered, increasing the quantity of excavated material considerably. More arti-
facts have since been excavated: as of 2004 the total numbers are 18 shields,
several decorated with typical incised circular lion processions, one dis-
playing a lion projecting from its center, exactly like those depicted on
the Musasir temple relief (noted above).99 Here is a list: 52 quivers, some
with their arrows inside; 16 helmets; two cauldrons (none with attach-
ments); 2 jars; 4 sikkatu, all inscribed; 174 spear heads; 931 arrows; and 28
gold rosettes, palmettes, and buttons. And only 15 % of the site has been
excavated. Another excavated object is a magnificent inscribed shield with
95 van Loon 1966, 8488, Salvini in Merhav 1991, 1213, and Belli in Merhav 1991, 2021,
gives the weights involved as many tons.

96 Kendall 1977, 28; Barnett 1982, 365.
97 In Merhav 1991, 13.
98 Belli in Merhav 1991, 21. One wonders what was the nationality of the art historian-
archaeologist present at Musasir.

99 ilingiroglu and Salvini 2001, 162163, figs. 11, 12, 21, 22.
646 chapter twenty
important iconographic information excavated at Upper Anzaf.100 We can

easily imagine what museum curators would have told us about the specific
loci of the Ayanis and Upper Anzaf shields had they been plundered and
then exhibited in their western museums: reputedly from, said to come
from, obviously from or is known to come from: the south shore of the
Caspian Sea, Malazgirt, Li, and maybe even Musasir.
An important contribution of Urartian metal artifacts is that many bear
royal inscriptions that identify and date them. Thus, a shield of Rusa (III?),
son of Erimena, was recovered at Toprakkale, suggesting that the site contin-
ued to function after the time of Rusa II.101 However, inscriptions do not nec-
essarily furnish a specific chronology for the terminal locus where they were
excavated. As Salvini wrote, at Karmir Blur and Toprakkale, both sites built
by Rusa II, earlier royal inscribed artifacts were recovered, and thus one can-
not date all artifacts recovered from these sites to the 7th century bc merely
because of the presence of 7th century inscriptions.102 And it is important
to be aware that no other ancient culture has as many inscribed bronzes as
does Urartu.103
Needless to argue, unexcavated inscriptions also are of value and cannot
be ignored. Although none of the inscribed metal artifacts in the Jerusalem
catalogue were excavated, they nonetheless disclose important information
about the beginnings of Urartian metal production: artifacts of Ishpuini
are manifest evidence that from the beginning of the kingdom in the late
9th century metal objects such as chariot and horse gear and vessels were
being made;104 the catalogue conveniently presents in an Appendix (pp. 355
357) all 39 inscribed objects in the exhibition. Bellis essay lists the many
types of artifacts that bear inscriptions (p. 48), also a list of 275 metal
objects known to bear royal names, from Sarduri I to Rusa III. Belli also
reports (p. 44) that 69 inscribed bronze artifacts are housed in Turkish
museumsof which only three derive from excavations. Recent excavations
have moderately increased this meager number; Belli knows of at least
220 additional plundered inscribed metal artifacts in foreign museums, but
does not know how many exist in private collections. This adds up to 290
100 Belli 1999, with relevant bibliography.

101 Salvini 1995, 110111, 114.
102 In Merhav 1991, 11.
103 Noted by Seidl 2004, 18. But she avoids directly noting in note 199 that not a single
example of an inscribed artifact assigned to Luristan has been excavated there, and they
are thus not heterogener Herkunft; see Muscarella 1988, 120, note 6.
104 Merhav 1991, 201, 217, 224, figs. 1.2, Nos. 17, 25, 27; see also Rehm 1997, 167169.
inscriptions on metal compared to about 500 on stone (now see Seidl 2004
for the most comprehensive listing).
There is one significant caveat that remains to be investigated. I have not
encountered anyone who has considered whether or not all the unexcavated
inscribed metal artifacts were in fact inscribed in ancient Urartu. No one can
deny for a second the absolute reality that a forger is able to skillfully carve
any Urartian inscription he chooses,105 all the more so a small one, a kings
name on an ancient metal artifactto enhance in one act its monetary
and cultural value to scholars and museum staffs (one envisages a smiling
forger imagining: think of all the royal Fundkomplexen they will recognize
and write about!). One could legitimately argue that until all unexcavated
inscribed inscriptions are testedbut not by a museum laboratoryit may
be impossible to argue that an ancient hand wrote the inscription. A warn-
ing is a good place to end a review filled with warnings.
Addendum
In May of 2005 I received U. Seidls study of Urartian bronze artifacts (2004,

see Bibliography). At this time I had completed this study, preparing to
send it to the publisher. I could therefore not give it full attention, but
note that her work and mine focus on essentially different studies. In a
few cases I incorporated some of her contributions and comments in my
manuscript. Much work was involved in Seidls creation. It is a major and
much needed contribution, the most comprehensive and viable study of
Urartian metal artifacts accomplished to date; it will serve as a handbook for
a long time. The author discusses the history of scholarship and a summary
of archaeological excavations and their metal artifact recoveries up to very
recent times (pp. 317). This is followed by a full discussion of the corpus
of inscribed artifacts, both from excavations and the antiquity market (18
124). Discrete chapters on a good number of categories of artifacts are given
full discussion and bibliographies: vessels, sculpture, candelabra, helmets,
shields, quivers, horse and chariot equipment, bells, belts, and plaques. Pins,
fibulae, jewelry, etc. are not discussed; one hopes a second volume will
complete the survey of the whole corpus.
Acknowledgements: I wish to thank Elizabeth Simpson and Mark Garri-
son for reading a manuscript of this paper and making valuable suggestions
and comments.
105 Or any inscriptionsee e.g. Muscarella 1988, 283284, note 4, and Muscarella 2000, 183,
No. 39.
648 chapter twenty
Bibliography
Amiran, R. 1966: Two Luristan Fibulae and an Urartian Ladle from Old Excavations
in Palestine, Iranica Antiqua 6, 8891.
Azarpay, G. 1968: Urartian Art and Artifacts: a Chronological Study (Berkeley).
Barnett, R.D. 1965: An Urartian Mirror. Anadolu Arastrmalar 2, 12, 5154.
Barnett, R.D. 1982: Urartu. In Cambridge Ancient History 3.2, 314371.
Belli, O. 1997: Der Bronzegurtel des Knigs Rusa in Malatya Museum. Anadolu 23
(Akurgal Festschrift volume), 293297.
Belli, O. 1999: The Anzaf Fortress and the Gods of Urartu (Istanbul).
Biscione, R. 1994: Recent Urartian Discoveries in Armenia: The Columbarium of
Erevan. Studi Miceni ed Egeo-anatolici 34, 115135.
Born, H. and Seidl, U. 1995: Schutzwaffen aus Assyrien und Urartu (Philipp von
Zabern).
Boysal, Y. 1961: Anzavurda Definecilerin Meydana ikardg: Eserleri. Belleten 25,
199212.
Calmeyer, P. 1986: Zu einem Ziselierten Bronzehelm des 8. Jahrhunderts. Archolo-
gische Mitteilungen aus Iran 19, 7986.
Caner, E. 1998: Bronzene Votivbleche von Giyimli (Rahden/Westf.).
ilingiroglu, A. and Salvini, M. (eds.) 2001: Ayanis 1 (Rome).
Collon, D. 1993: Review: Kellner 1991. Bulletin of the School of Oriental & African
Studies 56, 1, 125127.
Curtis, J. 1996: Urartian Bronze Belts. Zeitschrift fr Assyriologie 86, 118136.
Dezs, T. and Curtis, J. 1991: Assyrian Iron Helmets from Nimrud now in the British
Museum. Iraq 53, 105126.
Erzen, A. 1974: Giyimli Bronz Definesi ve Giyimli Kazisi. Belleten 38, 191213.
Garrison, M. 1994: Review of Merhav 1991. Ars Orientalis 24, 149151.
Ghirshman, R. 1964: Deux oreillers en bronze des rois dUrartu. Artibus Asiae 27,
4960.
Ghirshman, R. 1966: Deux bronzes des rois dUrartu. Artibus Asiae 28, 219222.
Gropp, G. 1981: Ein Pferdegeschirr und Streitwagen aus Urartu. Iranica Antiqua 16,
95123.
Haerinck, E. and Overlaet, B.J. 1984: Zur Funktion Einiger Uraratischer Bronzege-
genstande. Iranica Antiqua 19, 5370.
Hoffman, H. 1971: An Urartian decorated Bronze Strip from Diyarbakir. In D.G. Mit-
ten et alii (eds.), Studies Presented to George M.A. Hanfmann (Cambridge, MA),
6971.
Kellner, H.-J. 19751976 (= 1980): Ein datierter Silberfund aus Urartu, Anadolu 19, 56
67.
Kellner, H.-J. (ed.) 1976: Urartu: Ein Wiederentdeckter Rivale Assyriens (Mnchen).
Kellner, H.-J. 1977: Pectorale aus Urartu. Belleten 41, 481493.
Kellner, H.-J. 1979: Eisen in Urartu. In Akten des VII lnternationalen Kongresses fr
iranische Kunst und Archologie (Berlin), 151156.
Kellner, H.-J. 1980: Bemerkungen zu den Helmen in Urartu. Anadolu Arastrmalar
8, 207213.
Kellner, H.-J. 1982: Gedanken zu den Bronzen Blechvotiven in Urartu. Archologis-
che Mitteilungen aus Iran 15, 7995.
Kellner, H.-J. 1991: Grtelbleche aus Urartu. (Prhistorische Bronzefunde XII, 2) (Stutt-
gart).
Kellner, H.-J. 1993: Ein Assyrischer Beutehelm? In M. Mellink et alii (eds.), Aspects of
Art and Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet zg
(Ankara), 325328.
Kendall, T. 1977: Urartian Art in Boston: Two Bronze belts and a Mirror. Boston
Museum Bulletin 75, 2555.
Kohlmeyer, K. and Saherwala, G. 1984: Frhe Bergvlker in Armenien und im Kauka-
sus (Berlin).
Kroll, S. 1979: Urartu. Das Reich am Ararat (Hamburg).
Kroll, S. 1980: Urartu. Ein wiederentdeckter Rivale Assyriens (Muenster).
Kroll, S. 1997: Review: Salvini 1995. Die Welt des Orients 28, 203208.
Maass, M. 1987: Helme, Zugehr von Wagen und Pferdgeschirr aus Urartu. Archol-
ogische Mitteilungen aus Iran 20, 6592.
Merhav, R. 1981: A Glimpse into the Past. The Joseph Ternbach Collection (Jerusalem).
Merhav, R. (ed.) 1991: Urartu A Metalworking Center in the First Millennium BCE
(Jerusalem).
Merhav, R. 1995: Gold and Silver Pins from Urartu: A Typology and Methods of
Manufacture. Tel Aviv 21, 1, 129143.
Muscarella, O.W. 1978: Urartian Bells and Samos. Journal of the Ancient Near East
Society of Columbia University 10, 6172.
Muscarella, O.W. 1981: Giyimli Plaques. In O.W. Muscarella (ed.), Ladders to Heaven
(Toronto), 175178.
Muscarella, O.W. 1988: Bronze and Iron (Metropolitan Museum of Art).
Muscarella, O.W. 1992: Greek and Oriental Cauldron Attachments: A Review. In
G. Kpcke, I. Tokumaru (eds.), Greece between East and West: 10th8th Centuries
bc (Mainz), 1644.
Muscarella, O.W. 1997a: Review of Zimansky 1998. Bulletin of the Asia Institute 11,
205210.
Muscarella, O.W. 1997b: Review of Salvini 1995. Bibliotheca Orientalis 54, 5/6, 734
737.
Muscarella, O.W. 2000: The Lie Became Great (Styx).
Nagel, W. 19591960: Ein urartischer Helm aus dem Argisti-Magazin. Archiv fr
Orientforschung 19, 144147.
zgen, E. 1983: The Urartian Chariot Reconsidered: I. Anatolica 10, 111131.
zgen, E. 1984: The Urartian Chariot Reconsidered: II. Anatolica 11, 91154.
Piliposyan, A. 1996: An Urartian Bronze belt. Studi Micenei ed Egeo-Anatolici 37, 123
129.
Piotrovsky, B. 1970: Karmir Blur (Leningrad).
Rehm, E. 1997: Kykladen und Alter Orient (Karlsruhe).
Salvini, M. 1980: Iscrizioni cuneiformi Urartee su Oggetti di Metallo. Studi Micenei
ed Egeo-Anatolici 22, 181189.
Salvini, M. 1991: Una nuova iscrizione urartea. Orientalia 60, 4, 334346.
Salvini, M. 1995: Geschichte und Kultur der Urarter (Darmstadt).
Seidl, U. 1988: Urartu as a bronzeworking center. In J. Curtis (ed.), Bronzeworking
Centres of Western Asia c. 1000539BC (London), 169175.
Seidl, U. 2004: Bronzekunst Urartus (Mainz am Rhein).
650 chapter twenty
Sevin, V. and Kavakli, Ersin 1996: Van/Karagndz. Arkeoloji ve Sanat Yaynlar (An-
kara).
Simpson, E. 2005: Tall Tales: Connoisseurs, and The Fabrication of Archaeological
Context. Source 24, 2, 2841.
Tanabe, K. et alii 1982: Studies of the Urartian Bronze Objects from Japanese Collec-
tions (1). Bulletin of the Ancient Orient Museum 4 (Tokyo).
Tasyrek, O. 1973: The Karahasan Bronze Urartian Belt in the Adana Regional Mu-
seum. Anadolu 17, 197205.
Tasyrek, O. 1975a: Some Inscribed Urartian Bronze Armour. Iraq 27, 151155.
Tasyrek, O. 1975b: The Urartian Belts in the Adana Regional Museum (Adana).
Tasyrek, O. 1976: Urartian Figurines and Metal Vessels in the Adana Regional and
Gaziantepe Museums. Trk Arkeoloji Dergisi 23, 2, 106115.
Tasyrek, O. 1977: The Dedeli Bronze Urartian Belt in the Adana Regional Museum.
Trk Arkeoloji Dergisi 24, 1, 157160.
van Loon, M. 1966: Urartian Art (Istanbul).
Vanden Berghe, L. and de Meyer, L. 1983: Urartu een vergeten cultuur uit het bergland,
Armeni (Gent).
Yildirim, R. 1991: Urartian Belt fragments from Burmageit. In A. ilingiroglu and
D. French (eds.), Anatolian Iron Ages (Oxbow Monograph 13), 131148.
Zimansky, P. 1998: Ancient Ararat (New York).
PART TWO
ARTIFACTS, CULTURES, FORGERIES, AND PROVENIENCE

Section One
The Aegean and the Ancient Near East

chapter twenty-one
THE ARCHAEOLOGICAL EVIDENCE FOR RELATIONS

BETWEEN GREECE AND IRAN IN THE FIRST MILLENNIUM BC*
Over a period of many years objects of Iranian origin or reflecting Iranian

influences have been excavated in the Greek world. Classical archaeolo-
gists date their contexts between the late 8th and the 7th centuries bc, a
chronology not inconsistent with that assigned to the Iranian objects and
motifs in general. As more material from Greece and Iran is excavated (and
published), and as more research is devoted to old Greek excavations, schol-
ars concerned with Greek-Near Eastern cultural relations will perhaps be
able to recognize additional Iranian objects and their offspring in the West.
An important article published recently by Hans-Volkmar Herrmann has
demonstrated how careful re-examination of earlier site reports will yield
significant results in this area of study.1
The aims of this paper are both to review the conclusions of Herrmann
and other scholars who have discussed pre-Achaemenian Iranian and Greek
relations, and to offer a summary of what I regard as the evidence that should
be presented in determining what may and what may not be accepted
as examples of Iranian material occurring in the West. As we shall see,
the data indicates that Iranian goods and ideas reached the West in pre-
Achaemenian times, and that historians may legitimately cite their influ-
ences on Greek culture. Because of the subjective nature of some of the
conclusions presented here, this paper should be considered to be in the
nature of a preliminary report, as work in progress, and one that seeks to
generate future comments and refinements.
In the course of the Persian wars with Greece in the early fifth century bc
a large amount of personal and military equipment was abandoned by the
invaders and eagerly gathered by the victors. Ancient historians, Herodotus
* This article originally appeared as The Archaeological Evidence for Relations between
Greece and Iran in the First Millennium B.C., Journal of the Ancient Near East Society of
Columbia University 9 (1977): 3148.
1 Hans-Volkmar Herrmann, Frhgriechischer Pferdeschmuck Vom Luristantypus, JdI
83 (1968), 178. (I wish to thank Gunter Kopcke and Joan Mertens for making important
suggestions in the writing of this paper.)
656 chapter twenty-one
in particular, have furnished us with details of the abandonment and inven-

tories of the material collected by the Greeks. Thus, although surprisingly
few Persian objects have been discovered in Greece, we are faced with no
serious problems concerning which types of Achaemenian objects reached
Greece and the nature of their arrival.2 In pre-historic Greece, however, there
are no historical records to guide and instruct us, and archaeological activity
remains at present the only source of information concerning cultural con-
tacts between Greece and pre-Achaemenian Iran. Moreover, with respect to
the identification and boundaries of specific states then existing in Iran, and
their local cultural characteristics reflected in art, we know very little. Thus,
except for material obviously characteristic of Luristan in western Iran and,
to a certain extent, northwestern Iran, it is difficult to identify the exact area
within Iran where an Iranian object excavated in a foreign land may have
originated. Nevertheless, we are able to note that the material recovered in
the West is heterogeneous and that it is paralleled by objects excavated in
several areas within Iran, Luristan, northwestern Iran, the South Caspian
2 For a discussion of Achaemenid remains and influences in Greece see D.B. Thompson,
The Persian Spoils in Athens, in The Aegean and tbe Near East (New York, 1956), ed. Saul
Weinberg, 281291; Herbert Hoffmann, The Persian Origin of Attic Rhyta, AntK 4, 1 (1961), 21
26; The article by Anna Roes, Motifs iraniens dans L Art grec archaique er classique, Revue
Arch. IV (1934), 135154, juxtaposes too many motifs and cultures and presents conclusions
too broad to be of value. In BCH 85, 2 (1961), Chronique des Fouilles, 722 and pl. XXV, is
a report from Olympia of a bronze pointed helmet with an inscription identifying it as an
object captured from the Medesi.e. from the Persian army in the early 5th century bc (see
also my comment in AJA 73, 4 (1969), 479). And at least two Achaemenid gold lion bracteates
have been found on Greek soil, at Dodona and on Samothrace: BCH 80 (1956), 300, fig. 2;
Arcbaeological Reports 19651966, 19, fig. 33. J. Brker-Klhn in ZfA 61, 1 (1971), 138139, notes 31
and 32, fig. 15b, publishes a drawing of an apparently Achaemenid seal that was claimed
many years ago to have been found at Marathon: unfortunately there is no documentation
to support this claim. For a discussion of the lack of information regarding the find-spot
of the gold Achaemenid bracelet in the Karlsruhe Museum (said to be from Corinth),
see my Unexcavated Objects and Ancient Near Eastern Art, in Mountains and Lowlands
(Undena Press, California, 1977), ed. Louis D. Levine and T. Cuyler Young, Jr., 195. This latter
work includes a list of excavated Achaemenian Kleinkunst and their proveniences, 192196.
Andrew Oliver, Jr. in Persian Export Glass, JGS XII (1970), 9, believes that glass fragments
excavated at Olympia were imported from Persia , a conclusion yet unproven. For a
discussion of a 4th century quasi Achaemenid relief excavated in Athens see A.D.H. Bivar,
A Persian Monument at Athens, and its Connections With the Achaemenid State Seals, in
the W.B. Henning Memorial Volume, ed, M. Boyce, I. Gershevitch, (London, 1970), 4361. Bivar
concludes that the relief was sculpted by a Greekan opinion I share (at least we can state
the relief was not sculpted by an Achaemenid Persian.) That non-Persians manufactured
Achaernenid-style objects is proven by the Tomb of Petosiris reliefs from Egypt; see my
comments in BASOR 223, (October, 1976), 72; and Unexcavated Objects, 193f., No. 100.
relations between greece & iran in the first millennium bc 657
area, and western Iran in general. Obviously, only a judicious investiga-

tion of the art historical and archaeological background of all the material
under review will yield meaningful conclusions, and cautious, conservative
methodology must underlie the study of each piece.
Not surprisingly, the Iranian objects and motifs in the West are relatively
few when compared to those from North Syria, the Near Eastern area that
seems to have had the greatest artistic influence upon Greece. The Iranian
imports are also fewer than those from Phrygia, but apparently equal to
those from Assyria.3 Noteworthy, and of still undetermined significance, is
the fact that to date no recognizable Iranian object (as opposed to motifs)
has been recovered on the Greek mainland. Except for one piece discovered
on Crete, Iranian objects have been recovered only from Samos, an island
within sight of Asia Minor.
The example from Crete comes from Fortetsa and is the well-known
Luristan-type bronze open-work pendant depicting a master of animals
scene (Figure 1).4 From Samos there is the equally well-known bronze
spouted vessel at home in western and northwestern Iran (Figure 2),5 as well
3 For a discussion of North Syrian exports to the West, see my Near Eastern Bronzes in
the West: A Question of Origin, in S. Doehringer and David Mitten, eds., Art and Technology,
(Cambridge; Mass., 1970), 116 f. See also note 67 for Phrygian and Assyrian imports.
4 Berta Segal, Greece and Luristan, BMFA 41 (1943), 22 f., fig. 3; J.W. Brock, Fortetsa (Cam-
bridge, 1957), 199, pl. 114, unmistakable Luristan type; John Boardman, The Cretan Collec-
tion in Oxford (Oxford, 1961), 150, related to Persian (Luristan) bronzes that came via
North Syria; idemn., The Greeks Overseas (Penguin, 1964), 89; Roman Ghirshman, The Arts of
Ancient Iran (New York, 1964), 331, fig. 406; Herrmann, Pferdeschmuck, 26, fig. 20, luristanis-
chen Fassung; Pierre Amandry in Le Rayonnement des Civilisations Grecque et Romaine ,
Huitime Congrs International d Archologie Classique (Paris, 1963), 487, considers the Cre-
tan pendant to be one of the few objects srement iraniens found in the West. P.R.S. Moorey,
Catalogue of the Ancient Bronzes in the Ashmolean Museum (Oxford, 1971), 26, alone to my
knowledge rejects an Iranian attribution for this pendant: if anything [it is] Caucasian.
Note that Moorey claimed that Jacobsthal had changed his mind that this object was sub-
Luristan, but it is my impression that Jacobsthal was referring only to the Perachora buckle
(infra); see Paul Jacobsthal, Greek Prins (Oxford, 1956), 77, n. 1. Compare the Fortetsa pendant
to the same type of objects attributed often to Luristan or Iran, Andre Godard, Les Bronzes
du Luristan. (Paris, 1931), pl. 32, nos. 116, 117, pl. 35, nos. 148, 151; H. Potratz, Das Kampmotiv
in der Luristan-Kunst, Orientalia 21/1 (1952), pls. VIX, nos. 2335.
5 Cited many times, e.g., Segal, Greece and Luristan, fig. 2; Ghirshman, The Arts of
Ancient Iran, fig. 406; Herrmann, Pferdescbmuck, 26, fig. 20; Moorey, Catalogue, 276ff., 280;
idem, Ancient Persian Bronzes from the Island of Samos, Iran 7 (1974), 191, fig. 1 (hereafter
Moorey 1974); Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin,
1969), 101; for Iranian proveniences, Amandry, Le Rayonnement des Civilisations, 487; R.M.
Boehmer, Zur Datierung der Nekropole B von Tepe Sialk, AA 1965, 811f.: Judy Birmingham,
The Overland Route Across Anatolia, An. St. 11 (1961), 192; Boardman, Greeks Overseas, 89.
as the more recently excavated Luristan bronze standard finial fragment

(Figure 3), and a bronze figure of a mountain goat (Figure 4).6 These four
objects, which in Iran range in date from the late 9th or early 8th through
the 7th centuries bc, have generally been accepted as Iranian and their
identification present no problems.7
In addition to this primary listing of undisputed Iranian objects, there
is another category consisting of three types of objects, all from Samos,
whose identification as imports from Iran present no major difficulties. An
Iranian attribution for these objects, however, is slightly less demonstrable
for various reasons, and, therefore, I prefer to list them separately. Within
this category I place an open-work bronze bell (Figure 5), a bronze pendant
(Figure 6), and seven bronze goats, each set on a ring base (Figure 7).8
Open-work bells have been excavated not only in Iran but also in the
Caucasus, in Urartu, and in Europe (below). The particular type of bell
from Samos, of pomegranate shape, is commonly reported from clandestine
digging in Iran, but more significantly, it has been excavated in Iran at
Marlik, south of the Caspian Sea, as Sialk in western Iran, and at Hasanlu in
northwestern Iran, covering a period from ca. 1000 to 800bc.9 It seems to me,
6 Ulf Jantzen, gyptiscbe und orientalische Bronzen aus dem Heraion von Samos: Samos
VIII (Bonn, 1972), pl. 74, B896, pl. 72, B1282; for the latter see also Gnter Kopcke, Heraion
von Samos: Die Kampagnen 1961/1965 im Sdtemenos, Ath. Mitt. 83 (1968), 291f., fig. 33,
pl. 123. Birmingham, Overland Route, 192, mentions the Ibex at Samos as Iranian but gives
no clarification with regard to what object she refers; nor is there any further information
concerning the horse handle fitting and the heavily bossed harness and armour pieces from
Samos likewise attributed to Iran.
7 See for example the notes above and my review of Jantzen, Samos VIII in AJA 77 (1973),
237; Moorey, Catalogue, 280; idem 1974, 191, 193. Jantzen catalogued the Samos mountain goat
as Assyrian on pp. 70, 73; and see note 4 for Mooreys rejection of the Cretan pendant as
Iranian. Herrmann and J. Brker-Klhn in their reviews of Jantzen, Samos VIII, in Gnomon 47
(1975), 398, 399, and OLZ 70/6 (1975), 544, respectively, also accept the Cretan pendant and
the Samos mountain-goat as Iranian. Brker-Klhn, ibid., 540f., claims Samos B589 is Iranian,
an opinion I cannot accept.
8 Jantzen, Samos VIII, pl. 74, B1161, pl. 75, B1278, and pl. 58. Jantzen attributed the bell and
pendant to Iran, the goats to North Syria.

9 Muscarella, AJA 77 (1973), 237, Moorey, Catalogue, 138, and idem, (1974), 192, as well as
Herrmann, Gnomon 47 (1975), 398, and Brker-Klhn, OLZ 1975, 545, accept the Samos hell as
Iranian. For excavated examples, see E. Negahban, A Preliminary Report on Marlik Excavation
(Tehran, 1964), fig. 132; R. Ghirshman, Fouilles de Sialk (Paris, 1939), pl. LVI, S833; Hasanlu,
unpublished, 72.151, two examples. For bells said to be from Iran (all unexcavated except no. 12
from Sialk) see Jan Bouzek, Openwork bird-cage Bronzes, in J. Boardman & M.A. Brown,
eds., The European Community in Later Prehistory, Studies in Honor of C.F.C. Hawkes (London,
1971), 80, fig. 8. Bouzek does not cite the Marlik example and does not distinguish the Samos
pomegranate bell from the others; see pp. 88 and 94.
therefore, that we may with some confidence accept an Iranian source for
the Samos bell unless future excavations document a more widespread
distribution for the type.
The bronze pendant, unlike the bell, has to date no published parallels
from excavations in Iran or, to my knowledge, elsewhere in the Near East.
However, similar types of objects have been reported from clandestine dig-
ging in Iran, and on the basis of style an Iranian attribution for the pendant
is not precluded. Therefore, it is suggested that the pendant be accepted, at
least on a tentative basis, as an Iranian product, with the understanding that
we cannot accept proveniences offered by dealers as an historical reality.10
As for the goats, I have noted elsewhere that no excavated parallels are
presently known.11 Nevertheless, as with the pendant, the goats fit into
an Iranian background stylistically, and the few exact parallels known to
me are claimed to be from Iran (which without the stylistic underpinning
would be of doubtful value). On this basis, therefore, it is suggested that the
goats derived from Iran. It is of some interest to note that the goats were
excavated in five different areas at Samos and that each is slightly different
in height, width and body structure, which surely indicates that they should
be considered as seven separate objects, rather than as parts of a single unit.
These two groups, a total of thirteen objects, are to my mind the examples
that may legitimately be brought forward in discussions concerned with Ira-
nian material excavated in the West. If they are separated into two groups
it is because I am deliberately taking a conservative position in the conclu-
sion that the first category is more firmly groundedbecause derived from
excavationsin an Iranian background than the second; in the final analy-
sis the division may be more theoretical than factual.
Turning now to a direclty related subject, the recognition of specifically
Iranian motifs manifesting themselves in Greek art, we find ourselves in
a more difficult type of research. It is more difficult because the motifs
alleged to be Iranian have been translated into Greek forms and style and
are thereby at least one step removed from the original model. The problem
for those concerned with these orientalizing (Iranianizing) motifs is to
10 Muscarella, AJA 77 (1973), 237, Moorey 1974, 192, and Herrmann, Gnomon 47 (1975),
398, accept these objects, including the Samos example, as Iranian. For problems related to
proveniences offered by dealers, see my Ziwiye and Ziwiye: The Forgery of a Provenience,
JFA 4/2 (1977), 197219.
11 Muscarella, AJA 77 (1973), 236, where the two paraliel pieces are mentioned; Moorey
1974, 192 f., and n. 56, and Herrmann, Gnomon 47 (1975), 396f., also accept the Samos goats as
Iranian.
recognize in the Greek object or design the Iranian tradition by convincingly

matching it with the prototype preserved in its homeland. With the caveat
that subjective analysis plays more of a role in this type of research than that
with identifying actual imports, we may proceed.
To my mind there are at least a minimum of approximately four cate-
gories of objects within the Greek world that may be accepted with some
degree of certainty as having developed as a result of contact with Iranian
prototypes. Three of the categories have been systematically presented by
Herrmann in the paper mentioned above and therefore require only a brief
summary here. With regard to the first category, Herrmann has presented
good evidence and arguments for accepting an ultimately Iranian source
for the V-shaped horsebits excavated at Delphi (Figure 8) and Olympia.12 He
has demonstrated not only an iconographical but also a formal relationship
between the Greek examples and the V-shaped types that without doubt
derive from Iran (cf. Figure 9); he has also called attention to the rarity of
this shape among the known Greek horsebits. His conclusions are convinc-
ing and derserve acceptance.
The second category of objects consists of an animal or bird set on a shank
that is itself attached to a disc, examples of which have been excavated on
Rhodes (Figure 10) and at Samos.13 To my eyes the parallels cited by Herr-
mann as Iranian prototypes are not so close as one might wish to support
a Greek-Iranian relationship. However, aside from the many excellent par-
allels available on the art market and there attributed to Iran (Figure 11),
excavated examples are now known from Marlik where they functioned as
stamp seals.14 While it is not clear whether the Greek examples also func-
tioned as stamp seals, the formal similarities between these and the Iranian
types are too close to permit a rejection of the relationship suggested by
Herrmann.
12 Pferdeschmuck, 318, figs. 1, 4, 710, with parallels given in footnotes; P.R.S. Moorey,
Towards a Chronology for the Luristan Bronzes, Iran 9 (1971), 123, seems to support Cal-
meyers suggestion. See also J.A.H. Potratz, Die Pferdetrensen des Alten Orient (Rome, 1966),
figs. 32: f.g, 46: c, 60, pl. LVI, 133135.
13 Pferdeschmuck, 31 f., n. 117, fig. 26.
14 E. Negahban, The Seals of Marlik Tepe, JNES 36/2 (1977), 99f., figs. 1926; for unprove-
nienced examples attributed to Iran, see P.R.S. Moorey, Ancient Persian Bronzes in the Adam
Collection (London, 1974), 177 f., nos. 175179. The Caucasian example cited by Herrmann in
note 117 is not close enough; the Marlik examples were not available to Herrmann but give
excellent support for his conclusions. In part they also satisfy Roes suggestion that the Near
Eastern derivatives for the Greek examples would someday be recovered; Anna Roes, Pro-
tomes Doubles et Ttes d Animaux Gmines, Revue Arch. 35 (1932), 206.
The third category of objects consists of pottery vessels from East Greece
decorated with horizontal animal friezes and filler ornaments (Figure 12);
as I have argued elsewhere, they depend, ultimately at least, on an Iranian
source.15 Evidence for the Iranian origin of the animal frieze exists both on
excavated vessels from Marlik and on other vessels, which although clan-
destine finds, are stylistically surely Iranian (Figure 13). I am able to present
here another metal vessel, hitherto unpublished and in a private collection,
which in basic shape, construction, and general decorative scheme, is simi-
lar to the Metropolitan Museum example of Figure 13, and which probably
came from the same area in Iran (northwest?).16 (The neck part of this vessel
should be examined to see if any repairs have been made in modern times.)
The suggestion that the Greek animal frieze in horizontal zones may depend
on Iranian prototypes has recently received welcome support from Moorey.17
With regard to filler ornaments one must be on guard against isolating
minor details on Greek pottery and casually interpreting them as examples
of influence from one or another Near Eastern culture.18 However, it may not
be a mere coincidence that rosettes were commonly used as filler ornaments
on Greek pottery of the 8th and 7th centuries bc and that they were equally
popular on earlier and contemporary Iranian metal and terracotta vessels
15 O.W. Muscarella, A Bronze Vase from Iran and its Greek Connections, MMA Journal 5
(1972), 43 f.
16 Ibid. 49 f., n. 80, Heeramaneck collection [now in the Los Angeles County Museum of
Art]. I did not know the owner of the vessel when I wrote note 80. This vessel is also cited by
P.R.S. Moorey, Some Elaborately Decorated Bronze Quiver Plaques Made in Luristan, Iran
13 (1975), 25, n. 42. (N. B. In my article A Bronze Vase from Iran I inadvertantly neglected
to point out an important parallel for the metope type of decoration represented as a band
on the upper part of the bronze vessel and on other Iranian vessels [41f., note 23; fig. 9].
This very same type of decoration occurs on the fringe of the kilt worn by the boxer and
the archer on the Hasanlu gold bowl. Also note that the Objects Conservation Department
of the Metropolitan Museum of Art believes that the vessel represented in figures 1314 of the
bronze vase article is apparently modern: one of the motifs that convinced me it was ancient
was the metope type banded decoration! For this piece see also Oscar White Muscarella,
Unexcavated Objects, 172, no. 10).
17 Iran 12 (1974), 195. Note that Moorey in Iran 13 (1975), 25, considers both the Metropoli-
tan Museum and the Heeramaneck vessels to be laterl in date (Iron Age IIIB) than sug-
gested by me (ca. 1000800bc, i.e., Iron Il)which would make them closer in time to the
Greek examples I claim to be derivative; Peter Calmeyer, Reliefbronzen in babylonischem Stil
(Munich, 1973), 205, n. 442, c, on the other hand, dates the Metropolitan Museum vase earlier:
wohl zu spt datiert .
18 See also the comments of Herrmann, Pferdeschmuck, 33, n. 121; 34, n. 130; J. Weisner,
Zur orientalisierenden Periode der Mittelmeerkulturen, AA 1942, 392, n. 4, 437; Jack Benson,
Horse Bird and Man (Amherst, 1970), 67; Boardman, Greeks Overseas, 80.
(cf. Figures 13 and 14).19 Since the quarter-rosette has been attributed to Iran,
may we not assume the same source for the full rosette?20 Moreover, during
the same period bees were also represented on both Greek and iranian
vessels.21 Ought we to assume an independent, coincidental use of this motif
in both cultures, or can we hypothesize that the earlier Iranian examples
were the source for the Greek (mainly Cretan) representations? I do not find
it difficult to conclude that Greek artisans may have been made aware of
both rosettes and bees by the same Iranian vessels that were decorated with
animal friezes.
The fourth category that should be included in our discussions is a frag-
mentary bronze spout on which stands a horned animal in the round; it
was excavated at Aetos on Ithaca (Figure 15), and Herrmann argues that
the piece ist ohne Luristanvorbilder schwer denkbar.22 If a spout of this
type had been excavated in Iran no doubts as to its indigenous origin would
have been raised; in fact, a terracotta vessel with a very similar spout sup-
porting an animal was excavated in a late 9th century context at Hasanlu
(Figure 16).23 Given this striking parallel, I believe we may easily accept
Herrmanns conclusion that the Ithaca spout was an adaption or an Iranian
vessel shape. And in this context it should be remembered that it is a histor-
ical fact that Iranian vessels were capable of reaching the West (Figure 2).
One wonders whether the animals on both the Aetos and Hasanlu spouts
19 Ghirshman, Sialk, pl. 83: A D, 87: S1548, 90:9; Negahban, Marlik, figs. 109, 136, pl. XVI;
Muscarella, A Bronze Vase, figs. 1 f., L. vanden Berghe, Archologie de lIran Ancien (Leiden,
1969), pls. 107: c, 150: b, 171: a, 143: d. Compare also rosettes on pottery from Kultepe, T. zg,
Kltepe and its Vicinity (Ankara, 1971), pls. 17:3a, b; 21:1b; 22:6; on pp. 86f. they are dated to the
6th century bc, too late to relate them to the earliest Greek examples.
20 R.D. Barnett, Oriental Influences on Archaic Greece, in S. Weinberg, ed., The Aegean
and the Near East, 230; Ekrem Akurgal, The Art of Ancient Greece (New York, 1966), 194, won-
ders if the Greek rosettes came from Urartu because of their occurrence on siren attachments,
which to Akurgal are of Urartian origin, to me North Syrian; see Muscarella, Near Eastern
Bronzes, 110 f., for a summary of various opinions regarding their origin.
21 For a discussion of bees and birds, sometimes represented together with rosettes,
on Iranian vessels, see my A Bronze Vase, 42, notes 24 and 25 and Fig. 13 here. See also
T.J. Dunbabin, The Greeks and their Eastern Neighbors (London, 1957), 45, and Muscarella,
Phrygian Fibulae from Gordion (London, 1967), 60 f., where it is suggested that there might
be a direct connection between the geometric styles of Greece, Anatolia, and Iran: but this
is a complex topic that needs much research and thought before it is accepted as a definite
historical event, rather than as an hypothesis worthy of serious consideration.
22 Herrmann, Pferdschmuck, 30 f. By Luristan Herrmann means western Iran; 6, n. 26.
23 Ghirshman, The Arts of Ancient Iran, fig. 25; cf. also his fig. 102. Herrmanns parallels
cited in his n. 110 are not so obvious. The idea of an animal or bird perched on a spout occurs
also in Phrygia, E. Akurgal, Phrygiscbe Kunst (Ankara, 1955), pl. 24b, but the Hasanlu example
is obviously closer to the Aetos example than are the Phrygian examples.
served apotropaic roles and whether the role as well as the vessel form was
borrowed by the bronze workers of Aetos.
Two other types of objects excavated in Greece warrant special consider-
ation at this point of our discussion because the possibility that they devel-
oped from an Iranian source is still unresolved. The ambiguity exists because
these objects are not stylistically defined enough to allow one to come down
definitely on the side of those who argue that they are examples of Iranian
influence, or, equally, that they are not. Herrmann discerns Iranian influ-
ence in a fine bronze horsebit and on other objects that depict the master
of animals motif, in particular on a bronze horsebit from Messinia and ivory
and lead figurines from Sparta.24 The Iranian parallels he cites are obviously
pertinent, and ones first reaction is to accept them. But since the motif
occurs throughout the Near Eastwhence the Greek examples no doubt
derivethe master of animals motif per se must be excluded from con-
sideration as a specifically Iranian influence.25 At the same time, inasmuch
as the Messinia horsebit is V-shaped, it is by no means impossible that in
this instance the motif may also reflect an Iranian source. Thus, although
we should not include the Messinia horsebit within the corpus of definitely
Iranianizing material, we should not categorically exclude it.
One is equally tempted to agree with Herrmann in deriving from Iran the
many fenestered spheres surmounted by an animal or bird that have been
excavated in Greece.26 Examples of similar spheres are reported to come
not only from Iran but also from the Caucasus.27 Though it seems that the
24 Pferdeschmuck, 18 f., figs. 14, 15; cf. his figs. 11 and 12.
25 Moorey, Catalogue, 27. It is not clear if this motif was known in the Greek Bronze Age;
Benson, Horsee Bird and Man, 46 f., 147, n. 48. I leave out of discussion here any mention of the
Italian horsebits that have been related to Iranian prototypes (e.g. Herrmann, Pferdeschmuck,
13 f., notes 56, 57 [add T.J. Arne, Luristan and the West, ESA 9 (1934), 283]) both because
the issues of possible Italian-Greek-European relationships is outside my competence, and
because of Mooreys reservations, Catalogue, 27. Note also that I have consciously accepted
above the Cretan open-work pendant with a Master of Animals motif as Iranian (contra
Moorey) because of specific formal parallels of the whole object with examples considered
to be Iranian.
26 Pferdeschmuck, 32, with Greek, Iranian and Caucasian references in note 118. See also
Bouzek, Openwork, 87f., fig. 13.

27 Moorey, Adam Collection, 100 f., nos. 7175. Bouzek illustrares none attributed to Iran
and connects the Greek examples to those he calls Thrako-Cimmerian, unconvincingly to

my mind; his fig. 12:7 is without an excavated provenience. In A Note on Pre-Achaemenid
Bronze Standard-Tops from Western Iran, Iran 15 (1977), 143, Moorey states that similar
objects were excavated in Gilan, northwest Iran, but he cites the Adam Collection references
above. In 7000 Years of Iranian Art (Smithsonian Institution, 19641965), no. 422, a similar
object is said to come from Pir-Kouh, Gilan, but with no further references.
spheres attributed to Iran, some of which functioned as bells (cf. Figure 17),
appear to be closer to the Greek examples than those from the Caucasus, we
cannot know for sure which were available to the Greeks. Nevertheless, in
this instance it seems certain that either Iran or the Caucasusor both
was the source of inspiration; even if we cannot pinpoint a specific region,
at least we have evidence of influence reaching Greece from a general area.28
This brief discussion on the search for the oriental provenience of the
master of animals motif and the fenestered spheres leads to a related sub-
ject for study. Scholars are now becoming aware that it is sometimes diffi-
cult, if not impossible, to recognize in which specific culture of the Near East
a particular orientalizing motif or imported object may have originated, and
whence it came to the West. The evidence from controlled excavationsas
opposed to the unverifiable proveniences offered by dealershas demon-
strated that several types of objects were manufactured, or at least used,
at more than one near eastern region. Unless there is unambiguous evi-
dence, such as obvious stylistic indications, for one of these regions to be
indicated as the probable oriental source, it is methodologically safer to
conclude that the object had a general near eastern background, rather
than to guess at a specific source. We thereby avoid distorting the evidence
and assigning credit for cultural contributions to the West to one culture
when in fact it may actually belong to another.29 In the following passages
I shall comment upon some of those objects and motifs that to my mind
have been erroneously attributed solely to Iran, to the exclusion of other
regions.
We begin with a number of bronze objects excavated on Samos and
Rhodes, and called at different times mace heads, scepters, furniture fittings,
and cosmetic containers. (It is probable that these similarly shaped objects
28 Equally, if not more, difficult to place solely in Iran are the many bronze animal figurines
with suspension loops (Anhnger) excavated in the West. They have a wide distribution in
the Near East and the Caucasus; Herrmann, Pferdeschmuck, 31, nn. 111115; see also an example
from Hasanlu; M. Rad, A. Hakimi, The Description and Results of the Scientific Excavations of
Hasanlu. Solduz, (Teheran, 1960), fig. 1 opposite page 72 [in Persian]. It is also possible that
the addorsed heads of Samos VIII, pl. 73, B1130 may be accepted as Iranian-Caucasian in the
broad sense; Moorey 1974, 192, sees it as Caucasian, Muscarella 1973, 237, as possibly Iranian;
see also on the same page my comments regarding the open-work bird, BB762. For references
to material common both to Iran and the Caucasus, see F. Hancar, Kaukasus-Luristan, ESA
9 (1934), 47112.
29 For further discussion on near eastern proveniences and problems raised in attribution
see Muscarella, Near Eastern Bronzes, 109 f.; Herrmann, Pferdeschmuch, 33, and his Urartu
und Griechenland, Jdl 81 (1966), 79141.
were utilized for different functions).30 Judy Birmingham considered the

western examples to have been imported from Iran merely because she
encountered one in a Teheran dealers shop. But these objects have been
excavated not only in Iran at Hasanlu (Figure 18), but also in Cyprus, Assyria,
and North Syria; consequently, it is not possible to isolate one near eastern
source for the western imports to the exclusion of others, and to claim to
know where the Samos and Rhodes pieces came from. Moreover, stylistic
details characteristic of one culture are lacking, which means that we do
not know at present where they were manufactured; thus, the objects are to
be regarded as near eastern in general.31
The same problem exists with the various types of bells excavated on
Samos (Figure 19) because similar types are known from excavations in
Europe, the Caucasus, Iran, Urartu, North Syria, and Phrygia.32 One of the
Samos bells has been accepted on the basis of characteristic features as an
Iranian import (above), but the others from Samos, not so clearly differen-
tiated in style, must be accepted as deriving from some still undetermined
area or areas.
Herrmann has also argued that the fragment of a horsebit itself cast in
the form of a galloping horse, excavated on Rhodes, is an East Greek copy
30 Birmingham, Overland Routes, 192, 187, figs. 710; Calmeyer, Datierbare Bronzen, 91f.,
figs. 94, 95; Kopcke, Heraion von Samos, 294, pl. 126:4; R.H. Dyson, Jr., In The City of the
Golden Bowl, ILN (Sept. 12, 1964), 375, fig. 10; E. Gjerstad, Swedish Cyprus Expedition 4/2
(Stockholm, 1943), fig. 24, no. 11; R.D. Barnett, Layards Nimrud Bronzes and their Inscrip-
tions, Eretz Israel 8 (1967), 4 f., pl. VIII (called Syrian); Jantzen, Samos VIII, 56f., pl. 50 (called
North Syrian). Note that at Hasanlu at least one footed example contained kohl, the footed
end was sealed with a wooden plug, and a kohl stick was excavated juxtaposed, which means
that it could only be a cosmetic container; others found there are either not so clearly distin-
guished or are apparent mace heads.
31 For a discussion of proveniences and possible sources of manufacture, see Calmeyer,
Datierbare Bronzen, 91f. and the map of fig. 94, which should have the Cypriote and Hasanlu
examples added. See also Herrmann, Gnomon 47 (1975), 396.
32 Jantzen, Samos VIII, 81 f., pls. 79, 80; Hans Mbius, Kaukasische Glocken in Samos,
in E. Sprockhoff, ed., Marburger Studien, (Darmstadt, 1938), 150166, pls. 6669; Calmeyer,
Datierbare Bronzen, 111 f.; idem, Glocke, In RLA (Berlin, 1969), 427431. Mbius and Calmeyer
suggested that the Samos bells came from the Caucasus; Moorey, Catalogue, 138, suggested
they came either from the Caucasus or Urartu; Bouzek, Openwork, 88, sees the Samos bells
related more directly to Caucasian bells but there need not be any direct connection;
Herrmann, Pferdeschmuck, 31, recognized the bells wide distribution. For other near eastern
bells see Guitty Azarpay, Urartian Art and Artifacts (California, 1968), 25, fig. 6; L. vanden
Berghe, La Ncropole de Khurvin (Istanbul, 1964), pl. 37, also pl. 29, no. 216; R.H. Dyson, Jr.
Hasanlu and the Solduz and Ushnu Valleys , Archaeologia Viva 1 (1968), 90, upper left;
Potratz, Die Pferdetrensen, 161, n. 1; Gordion, unpublished, 1910 B300, from the destroyed
Phrygian level.
of a Luristan bit,33 even though he was aware that the closest parallels are
an electrum example from Nimrud and several representations depicted
on Assyrian reliefs. To Herrmann, both the Rhodes and Nimrud examples
were derived from Luristan. Based on stylistic grounds, expecially the lack
of a ground line and the galloping position of the horse, Calmeyer and
Moorey have rejected an Iranian attribution for the Lindos bit, and consider,
correctly I believe, that it is Assyrian and came from there.34 This conclusion
obtains equally for the similar bits that were excavated on Samos.35 That
Iranian horsebits were probably the inspiration for the Assyrian examples
is not to be denied, but this is not at issue here: the horsebits from Samos
and Rhodes came from an Assyrian workshop and they cannot be cited as
examples of Iranian imports in the West.
In my review of Samos VIII I stated that a fragment of a bronze vessel
excavated on Samos was to be considered an Iranian import.36 I now real-
ize that although very similar vessels have been excavated at War Kabud
and at Tepe Guran in western Iran, other examples have been excavated at
Uruk in Mesopotamia.37 Most of the known examples of this bronze ves-
sel shape, those excavated and those from dealers shops (Figure 20), are
reported to come from Iran (which is why I originally attributed the Samos
vessel there), but the excavated examples from Uruk must also be intro-
duced into the discussion. At this time we do not know whether the Uruk
vessels were Iranian imports into Mesopotamia, although I suggest that
this conclusion is probable. In any event, we do not know at present if
the vessel came from an Iranian or from a Mesopotamian shop, and it is
33 Pferdschmuck, 22 f., fig. 17; 13, n. 53. In Gnomon 47 (1975), 397, Herrmann changed his
mind and sees the Samos bit as a local copy of an Assyrian bit.
34 Calmeyer, Datierbare Bronze, 114; Moorey, Iran 9 (1971), 123f.; idem., Iran 2 (1964), 194;
also Brker-Klhn, OLZ 1975, 540.

35 Muscarella, AJA 77 (1973), 236; Jantzen, Samos VIII, 64f., pl. 61, there called North Syrian.
36 AJA 77 (1973), 237; see also Calmeyer, Datierbare Bronzen, 115f., Moorey, Iran 13 (1974),
194. I believe that Mooreys fig. 4 is not correctly drawn as there should be a sharper carination
at mid-point; cf. Kopcke, Heraion von Samos, pl. 127:1.
37 L. vanden Berghe, La Nekropole de War Kabud, Archaeologia 18 (1967), 60f.; J. Meld-
gaard et al., Excavations at Tepe Guran, Luristan, Acta Archaeologica 1964, fig. 30. Compare
the War Kabud vessels to Eva Strommenger, Gefsse aus Uruk (Berlin, 1967), pl. 32, espe-
cially no. 78. Of some interest is the juxtaposition of the spouted vessel and and carinated
example from Tepe Guran, both types of which have been excavated at Samos.
I would also now claim that the knobbed mace from Samos (Jantzen, Samos VIII, pl. 51,
B574: upside down?), which I suggested in AJA 77 (1973), 236, has good parallels in Iran,
also has good parallels elsewhere and therefore might better be labelled near eastern; see
Calmeyer, Datierbare Bronzen, 108.
therefore a safer conclusion to suggest that it be considered a near eastern

import, with no sharper distinction.
Another statement made in the Samos VIII review must be modified;
it concerns the near eastern origin of a bronze horsebit terminating in
animal heads excavated on Samos. I suggested there38 that the horsebit B951
is probably Iranian as bits of this type occur there , but neglected to
note that similar bits, with animal-headed terminals, are also known from
the Caucasus.39 It would seem, therefore, that the Samos bit could have
come either from Iran or from the Caucasussimilar in this respect to the
fenestered spheres discussed above. A stylistic examination of the animal
heads on the Samos bitnot clearly seen in the published photographs
might help scholars to be more specific in the future.
A bronze strip from Fortetsa on Crete was compared by its excavator to
Luristan belts, an opinion supported by Herrmann.40 I take issue with these
opinions because it appears that the parallel suggested is too general and
is based on comparing belt to belt, disregarding stylistic references. Peoples
from more than one culture of the Near East wore decorated belts,41 and it
is not possible to isolate one region to the exclusion of others as the source
for the Fortetsa stripwhich may in fact not have been a belt.
Equally tenuous is Ghirshmans attempt to derive the terracotta boots
excavated in Greek tombs from Iranian models.42 Terracotta and bronze
boots are known from Iran but they also occur in Urartu and the Caucasus
in the first millennium bc. Moreover, and significantly (below), they are
documented both in Anatolia and Greece in the second millennium bc.43
Only by ignoring all the evidence could an Iranian model for the Greek
terracotta boots be claimed.44
38 AJA 77 (1973), 236.

39 Potratz, Pferdetrensen, figs. 48, 94, 95; Moorey 1974, 194; Brker-Klhn, OLZ 1975, 540.
40 J.K. Brock, Fortetsa (Cambridge, 1957), no. 1568, pls. 115, 168; Herrmann, Pferdeschmuck,
30.
41 Boardman, Greeks Overseas, 107 f.; P.R.S. Moorey, Some Ancient Metal Belts: their
Antecedants and Relations, Iran 5 (1967), 83 f.

42 Arts of Ancient Iran, 336.
43 Guitty Azarpay, Two Urartian Boot-Shaped Vessels, Artibus Asiae 27 (1964), 6171, for
a discussion of these objects in Greece and the Near East; see also vanden Berghe, Khurvin,
63, pl. 27; zg, Kltepe and its Vicinity, 101, pl. XXXIV.
44 Ghirshmans attempt, Arts of Ancient Iran, 341, to link bucchero pottery in Italy with grey
wares in Iran, and winged creatures in Greece to those from Iran, are equally unacceptable
and only further complicate the issue under study here. However, his discussion on p. 343
regarding the use of pins as votive offerings in Greece and Iran is probably correct, except
that this idea was prevalent all over the Near East, at least with regard to fibulae, see Oscar
White Muscarella, Fibulae Represented on Sculpture, JNES 26 (1967) 84f.
A few other objects need only be mentioned here for they have all been
discussed by Moorey who rejects the Iranian background alleged by others.
The best known of these is the bronze buckle excavated at Perachora
in Greece by Humfry Payne. Moorey believes that the piece actually has
affinities with the Caucasus; it seems certainly to be near eastern and, if
not necessarily Iranian, then possibly Caucasian.45 Moorey also rejected as
Iranian two fine bronzes excavated on Samos. One (B1211) consists of two
confronted equids, and is considered by him to be near eastern, with no
specific source suggested: the other (B1130) consists of addorsed wolf heads
each devouring a bull head, and is assigned to the Caucasus. While neither
object would cause surprise if it were excavated in Iran, one could agree
with Moorey that an Iranian attribution, to the exclusion of other regions, is
not compelling, and, therefore, I tend to support his position.46 In addition,
Moorey has correctly denied an Iranian attribution to a bronze object from
Olympia, which Herrmann (apparently unknown at the time to Moorey)
had implicitly claimed to be related to a mirror handle said to be from
Luristan.47
Whereas after critical analysis an Iranian attribution might be rejected
for the objects just discussed, most, if not all, have at least certain features
that allow such attribution to be understandable. This is not the case with a
particular motif that has been presented as an example of Iranian influence
in the West, and for which the evidence offered is nonexistent. Ghirshman
has cited the mounted archers employing the Parthian Shot on one of
the well-known Cretan bronze shields and maintained that the motif can
only have come from Iran.48 However, to my knowledge this motif is not
45 Moorey, Catalogue, 26, nn. 4, 5; Herrmann, Pferdeschmuck, 27, also rejected an Iranian
attribute claiming that its origin is not clear (but cf. Gnomon 47 [1975], 399); Boardman,
Greeks Overseas, 89, thought that it may have come from Iran; Amandry, Le Ravonnement
des Civilisations, 487, believed it to be Iranian.
46 Jantzen, Samos VIII, 74 f., Luristan, pl. 75; Moorey, Iran 12 (1974), 191f.; cf. Muscarella AJA
77 (1973), 237, they could tentatively be considered Iranian, which is still true although
modified by the Caucasian factor. Herrmann, Gnomon 47 (1975), 398f., also rejected an
Iranian attribution.
47 Moorey, Catalogue, 26; Herrmann, Pferdeschmuck, 30.
48 Arts of Ancient Iran, 341, fig. 432. A bronze bowl in the Ashmolean Museum that
depicts the Parthian Shot is to my mind incorrectly listed as having derived from Olympia:
F. Studniczka, Jdl 22 (1907), 165; H.T. Bossert, Altsyrien (Tbingen, 1951), 803; Dunbabin,
Greeks, pl. VII, 1. In fact, the bowl was purchased by the Ashmolean in 1903 and attributed
by the dealer to Olympia, which means that it cannot be legitimately cited as an example of
an import to Olympia or to any other site (information about the acquisition supplied to me
by P.R.S. Moorey).
depicted in Iranian art before the Sasanian period, but it is depicted in the
earlier art of Assyria, Urartu, and on Cypro-Phoenician bowls.49 Further, it is
mentioned as an Urartian technique of fighting by Sargon II of Assyria in his
report on his eighth campaign in 714 bc.50 To relate the representations on
the Cretan shields specifically to Iran, where to date there is no evidence for
its depiction in the first millennium, and to ignore the multiple near eastern
examples, is a distortion of the facts.
We come now to the final category of objects to be discussed in the
context of possible Greek-Iranian contacts. Instances occur where a motif
or object excavated in a first millennium bc site in Greece has parallels in
the contemporary Near East as well as in the earlier second millennium bc
Greek Bronze Age. In this connection we may cite, besides the terracotta
boots mentioned above, double-headed animals, two animals sharing one
head, animal-headed pins, and a bird mounted on the back of a horse, and
possibly still others. These motifs are represented in Greece both in the
round and painted on pottery;51 all have been cited by some scholars as
examples of Iranian influences on Greek artimproperly, it will be argued
here.
In these cases it is difficult to determine whether we are witnessing
the reappearance of motifs locally preserved, recognized and deliberately
developed from the earlier period, or whether in fact we are witnessing a
49 T. Sulimirski, Scythian Antiquities in Western Asia, Artibus Asiae 17 (1954), 290f.,
fig. 1. Sulimirski and others, e.g. R.H. Dyson, Jr., Problems of Protohistoric Iran as Seen from
Hasanlu, JNES 24 (1965), 208, and T.C. Young, Jr., The Iranian Migration into the Zagros,
Iran 5 (1967), 20, have assumed that the warriors on this relief are Scythians, a conclusion
I consider to be dangerous. There are no identifying inscriptions on the relief, which was
sculpted before the time we first learn of Scythians in the Near East; see also K. Jettmar, Art of
the Steppes (New York, 1964), fig. 44. R.D. Barnett, Assyria and Iran , Survey of Persian Art
14 (1967), 2997, sees the figures on this relief as Iranian, inferring that they are Zamuans, to
my mind not a properbecause it is a guessconclusion.
For other examples of the Parthian Shot depicted in art, see Eva Strommenger, 5000 Years
of the Art of Mesopotamia (New York, 1964), figs. 242, 243, practiced by Arabs; Potratz, Pfer-
detrensen, Pls. IX:15, XXXIV:7; D. Randall Maciver, Villanovans and Early Etruscans (Oxford,
1924), pl. 38: 1, 2; M. Rostovzeff, The Parthian Shot, AJA 47 (1943), 180f., n. 15. In Urartu it
is represented on a belt in the Ashmolean Museum and on another in the Norbert Schim-
mel Collection; Oscar White Muscarella, ed., Ancient Art, The Norbert Schimmel Collection
(Mainz, 1974), no. 133, panel 20. Note that the archer on the Cretan shield is represented rid-
ing backward in the saddle, which also occurs in Sasanian art; R, Ghirshman, Parthians and
Sassanians (London, 1962), figs. 248, 250, 251.
50 D.D. Luckenbill, Ancient Records of Assyria and Babylonia (Chicago, 1927), 2, para-
graph 158, describes Urartian horsemen turning around in the saddle when fighting.
51 Segal, Greece and Luristan, 72 f.; Herrmann, Pferdeschmuck, 33f.
new appearance, solely derived from recent stimuli from the East. Were the
motifs copied by Greek artists of the first millennium bc who discovered
them on their own soil as a consequence of an interest in their Bronze
Age heritage, or were the motifs observed and adapted because they were
first noticed on oriental objects? Or, to mention yet another possibility, did
the Greek artists discover a motif as a result of knowledge from their own
past but used it in art only after exposure to oriental exotica? This last
suggestion seems too elusive to document without textual evidence and
may actually beg the question. For we would actually be arguing that the
oriental stimulus was the primary source and the historically significant
one; cultural continuity would have played no creative role, and the term
orientalizing would be the correct term to describe the action.
Thus, when Herrmann suggests that the double-headed animal, which he
recognized had a Bronze Age history both in the Near East, and in Greece,
re-emerged in late Geometric Greek art because of contemporary influence
from Iran (der Zweiten Rezeption), he is actually arguing that the motif
is an example of oriental influence, not of cultural continuity.52 The Greek
examples of this motif (Figure 21) appear close in form to contemporary
examples attributed to Iran (Figure 22), but it is argued here, pace Her-
rmann, that the issue of cultural continuity cannot be set aside; it must be
an integral part of the discussion. To my mind, once it has been established
that a given motif used in first millennium bc Greek art had both an earlier
local history and a contemporary life in the Near East, we cannot presume
that the latter occurrence was the stimulating force. To accept as fact either
that the oriental source was the cause for the Greek artists use of a motif,
which was already known to him, or that it was cultural continuity that was
the cause, is in the final analysis sophisticated guesswork. Therefore, this
charged category of material should be kept separate in our minds subject
52 Pferdeschmuck, 34 f., nn. 130131; cf. also Benson, Horse Bird and Man, 70, who believes
that the animal frieze was known to 8th and 7th century bc Greeks from their Mycenaean
heritage but that they decided to use them in art only after they were exposed to oriental
models; i.e., that the motif is an orientalizing one according to the position taken here. To
the three possible explanations suggested for the use of a motif in the first millennium bc
one might add another: thc motif was independently borrowed from the Near East twice,
once in the Bronze Age, the second time centuries later (an idea suggested by Roes, Revue
Arch. 4 [1932], 198). This theory, however, also rejects continuity and actually argues for the
new arrival in the first millennium of the motif, i.e. that it is an example of orientalising. For
evidence of first miliennium bc physical contacts with Mycenaean remains, see A.M. Snod-
grass, The Dark Ages of Greece (Edinburgh, 1971), 394 f., and Benson, Horse Bird and Man, 109f.,
115 f., 121 f.
to on-going research; we cannot rise above the evidence by imposing sub-

jective concepts on the inadequate data. In short, we may ask the ques-
tion about possible sources, but we do not know which stimulus set off the
spark.
As with the double-headed animal, the motif of two animal bodies shar-
ing one head en face (both motifs may be related) had a long history in
ancient art. It too may be seen represented in the art of both the Greek world
and the Near East in the second millennium bc, and it was drawn on metal
and pottery in Greece and Crete in the first millennium bc.53 Barnett and
Ghirshman, ignoring the Greek Bronze Age evidence, claim that its occur-
rence in Greece in the first millennium was a result of cultural contacts with
Iran.54 The motif apparently does exist in Iran in the first millennium bc
although its occurrence is not so easily documented as alleged because it is
represented to date only on objects that derive from dealers shops. Further,
one of the alleged Iranian parallels cited as a source for the Greek exam-
ples is to my mind of doubtful authenticity.55 In any event, Iran was not the
first nor the only region to have depicted this motif in art, and given its local
Bronze Age history in Greece, the cultural source for the first millennium bc
Greeks eludes us.
Equally difficult to pin down to a specific source are the first millennium
bc Greek examples of pins with animal finials. They occur not only in the
Caucasus and in Iran at this time,56 but also in Greece and the Near East in
53 Pierre Amiet, La Glyptique Mesopotomienne Archaique (Paris, 1961), pl. 26, 425, pl. 66:
890; Ernst Herzfeld, Iran in the Ancient East (London, 1941), 163, fig. 278, below; Arthur Evans,
The Palace of Minos IV 2 (London, 1935), figs. 575577; Humfry Payne, Necrocorinthia (Oxford,
1931), 52, fig. 12, pl. 16:4; Muscarella, Norbert Schimmel Collection, no. 15; Ghirshman, Arts of
Ancient Iran, figs. 581, 582.
54 Oriental Influences, 232, pl. XXI: 1, 2, 3; Arts of Ancient Iran, 316, figs. 383, 384, and page 335
(see note 55).

55 Oriental Influences; pl. XXI, 2; Arts of Ancient Iran, fig. 383. Figure 384 appears to be
genuine; I know of another example of this motif on a disc pin of first millennium date in
a private collection that seems to be genuine; see also Muscarella, Unexcavated Objects, 184,
no. 156. For an Achaemenian example of this motif, see John Boardman, Pyramid Stamp Seals
in the Persian Empire, Iran 8 (1970), 35, fig. 12. The gold appliques in 7000 Years of Iranian
Art, no. 456, and Kunstschtze aus Iran (Zurich, 1967), no. 868, illustration no. 18a, may be
genuinebut I have not examined them personally.
56 For excavated examples in Iran, see Ghirshman, Sialk, pl. XXIX, 1, top; Negahban, Marlik,
fig. 131 (cf. fig. 85); at ieast one example, with a rams head, comes from Hasanlu, unpublished:
7050. For others attributed to Iran, see Andr Godard, Les Bronzes du Luristan (Paris, 1931),
pl. XXXII; Moorey, Catalogue, 191 f., 197 f., pls. 50, 53; Herzfeld, Iran in the Ancient East, 154f.,
fig. 275, pl. XXXI; for the Caucasus, see F. Hanfcar, Die Nadelformen des Kaukasusgebietes,
ESA 7 (1932), figs. 10, 17, 18.
the second millennium bc.57 Again given these facts, Herrmanns suggestion
that an Iranian source sufficiently explains the motif in Greece must be
rejected.
Years ago Berta Segal suggested that the representation of a bird on a
horses back was dependent on Iranian prototypes; the motif occurs in
first millennium Greek art in the round and on pottery.58 Both Benson and
Herrmann have rejected Segals conclusions, the former because he believes
that the motif developed within Greece itself and is an example of cultural
continuity from the Mycenaean period, the latter apparently because it is
too causal a motif.59 There is yet another reason for rejecting the suggested
Greek-Iran relationshipnamely, that in the Near East Iran was not alone
in depicting the motif in art.60
How did the objects and motifs arrive in Greece? It could indeed be argued
that suggestions concerning the routes by which the material travelled from
the various areas within Iran can only be speculative.61 Yet the possible
routes available to us for serious consideration are not so many and may be
discussed with some profit; it is even possible to point to a specific, known
route as probably the one on which goods moved from East to West.
57 Herrmann, Pferdeschmuck, 32, n. 119; Moorey, Catalogue, 191f.

58 Segal, Greece and Luristan. 72 f.
59 Benson, Horse Bird and Man, 142, n. 71, 143, n. 80; Herrmann, Pferdeschmuck, 33, n. 121;
for a recently discovered example of this motif, see Marvin H. Pope, The Scene on the
Drinking Mug from Ugarit, in H. Goedicke, ed., Near Eastern Studies in Honor of W.F. Albright
(Baltimore, 1971), 395, fig. 2. Compare also Pierre Amandrys study of the widespread motif
of the reclining goat, Un Motif Scythe en Iran et en Grce, JNES 24/3 (1965), 149160.
Amandry rejects a specific near eastern background for the Greek use of this motif and raises
the questions (without attempting to answer them) whether it was an independent use in
the first millennium or whether it resulted from continuity from the Mycenaean period.
Benson, Horse Bird and Man, 58, 150, n. 92, cites the motif as an example of a true Mycenaean
renascence.
60 Muscarella, Near Eastern Bronzes, 120, fig. 11, North Syrian style; see also Segal, Greece
and Luristan, fig. 121; Herzfeld, Iran in the Ancient East, fig. 286; Pope, The Scene on the Drinking
Mug, fig. 2.
61 Here we may note the statement of M.I. Finley, Early Greece: The Bronze and Archaic
Ages (New York, 1970), 39, Archaeology alone, it cannot be repeated often enough, rarely can
reveal the mechanism of foreign relations even when it unearths great quantities of foreign or
foreign-inspired goods. And Dunbabin, Greeks, 15, speaking about historical interpretation of
archaeological material, says: The historian who is to attempt to use archaeological material
must have a clear idea how much to expect and how to go about it a superficial treatment
will lead to many pitfalls, mistakes of method, misunderstandings and misconceptions,
errors of detail and errors of scope, and particularly, ignorance of what is possible and of
the limitations on what archaeological evidence may be used to establish.
While there can be little doubt that central Anatolian goods travelled
overland across Anatolia to the West in the 8th century bc,62 there is no
evidence to suggest that goods originating in western and northwestern Iran
also travelled by this route. And there are two archaeological indications
that suggest that Iranian goods did not travel by this route. The first is
negative, but significant: It seems clear on the basis of recent research that
Urartian goods did not travel West across Anatolia, nor, it would seem, did
they travel West by any route.63 Thus, given our present knowledge, there
is no reason to assume that the overland Anatolian route extended as far
east as Urartu. Following upon this, how then may we hypothesize that
goods originating in Iran moved north through Urartu in order to cross
Anatolia by a route which seems not to have existed? We cannot; and if
the northern route is rejected, we must equally reject an alleged Black Sea
route to the West.64 We must look elsewhere, and this brings us to the second
archaeological indication mentioned above.
In the 9th century bc a vigorous trade existed between northwestern Iran
and Assyria and North Syria. The evidence consists of a large quantity of
imported Assyrian and North Syrian objects excavated at Hasanlu, many
of which are still unpublished.65 The existence of these imported objects
conclusively documents an overland route connecting North Syria, and ulti-
mately the Mediterranean Sea, on one end, and Assyria and Northwestern
Iran across the Zagros passes on the other. Whether this route was used in
the 8th and 7th centuries is, of course, not proven. But inasmuch as Assyria
continued to be interested in Iran during this period and was still a power
from the late 8th through most of the 7th century, it is probable that the
route was not discontinued. To summarize, we have archaeological evidence
62 For a summary of routes suggested for the movement of material over long distances
within the Near East to Greece, see Birmingham, Overland Routes, 185f.; Bouzek, Openwork,
94, dismisses an overland route from the Caucasus across Anatolia.
63 For evidence suggesting that no Urartian material reached the West and a rejection
of conclusions to the contrary, see Hans-Volkmar Herrmann, Urartu und Griechenland, Jdl
81 (1966), 79141, and Muscarella, Near Eastern Bronzes, 111f., 122. Recently S. Salvadori, An
Urartian Bronze Strip in a Private Collection, East and West 26, 1/2 (1976), 106, ignoring the
recent literature and relying on Pallottino, still maintains that Urartian goods reached the
West in quantity.
64 This route was championed by Barnett, Oriental Influences, 228.
65 For published information on Assyrian and North Syrian contacts with Hasanlu, see
Dyson, Problems of Protohistoric Iran, 199, and Muscarella, Hasanlu 1964, BMMA 25/3 (1966),
121135. The subject is more fully explored in my forthcoming A Catalogue of tbe Hasanlu
Ivories.
for only one general route by which Iranian goods apparently reached the
West: across the Zagros Mountains via Assyria and North Syria.
Related to our concern regarding trade routes is our interest in discov-
ering which groups or categories of people transported the material to the
West, and why. Was some of the material carried long distance from Iran
by an Iranian trading group, or were they carried in stages by groups of for-
eign merchants? Was some of the material transmitted from the emporia on
the North Syrian coast to Greek cities by Greek merchants, or foreign mid-
dlemen, or both, and did the objects arrive as items of trade? Were some
of the objects brought to the West by Greek or near eastern merchants, or
travellers, or pilgrims as votive objects?66
Leaving the question of motifs aside, we may hypothesize that some of
the objects discussed above had a votive value while others seem to have
been secular in functionalthough, it must be stressed, these distinction
are highly subjective. Of the supposed votive objects we might include the
open-work Luristan type pendant from Crete (Figure 1), the bronze Luristan
finial (Figure 3), and perhaps the mountain goat (Figure 4) and the seven
goats (Figure 7) from Samos. Of the apparent secular objects there are
the spouted vessel (Figure 2), the bronze pendant (Figure 6), and the bell
(Figure 5), all from Samos, But even if our tentative distinctions between
votive and secular objects are correctwhich is by no means certain, as
we cannot claim to know how each object was used or whether a given
object had different values under different conditionswe are ultimately
incapable of reaching conclusions concerning who brought each object. For
it could be argued on the one hand that merchants bring secular goods and
individuals (not necessarily merchants) bring votive goods; and on the other
hand that merchants alone may bring secular objects to sell and votive ones
to dedicate in order to secure personal safety and good trade.
66 Note the views of John Boardman, R.D. Barnett, T.J. Dunbabin, and W.L. Brown that
some of the oriental objects were made in the West by near eastern craftsmen; see Muscarella,
Near Eastern Bronzes, note 9; for Boardman see also Orientalien auf Kreta, in Dadalische
Kunst auf Kreta im 7. Jahrhundert (Hamburg, 1970), 1425, and The Khaniale Tekke Tombs, II,
BSA 62 (1967), 63 f. My aim in note 9 was not to deny the recognized value and importance for
scholars to know if near eastern craftsmen actually migrated to the West and worked there.
Rather it attempts to state that, given the fact that we do not know if migration of craftsmen
occurred, in the final analysis we are forced to concern ourselves with what we have, which
are the objects themselves. Are they orientalimports whether made locally or abroad
and are they orientalizing, resulting from an imported idea or motif adapted to local style
and needs? At present, I see no need to speculate about travelling Iranian craftsmen.
Herodotus has informed us of extensive gifts sent as dedications by near

eastern kings to various Greek sanctuaries (I:14, 50, 61, 52, 92) beginning
in the late 8th century bc with King Midas of Phrygia and followed by
the Lydian kings Cyges and Croesus. There may have been other kings
not mentioned by Herodotus who sent dedications, or the three he does
mention may have been the only ones.67 In any event, it would be premature
to suggest that some of the many Iranian monarchs ruling the various states
in the pre-Achaemenian period sent royal dedications to Greece. In short,
the questions concerning which groups brought the objects to the West, and
how and why they came, must at present remain unanswered.
Another question arises, a tantalizing one, and one also not possible to
answer: Did the Greeks know that the Iranian objects and motifs came from
beyond the Zagros Mountains, or were they considered to be simply more
examples of exotic material that arrived from somewhere in the East? If we
knew the answers to some of the other questions asked, we would be closer
to an answer for this one.
67 In addition to the throne sent to Delphi by King Midas (Herodotus I:14), other objects
of Phrygian origin that reached the West are as follows: from Samos there are three fibu-
lae, five belt buckles, possibly a belt fragment, apparently two bowl handles with spools,
and some sherds. Phrygian fibulae have also been excavated at Olympia, Argos, Perachora,
Tegea, Sparta, Lindos on Rhodes, and from Lesbos; bronze omphaloi have been excavated
at Olympia, Argos, Perachora, Delphi, Lindos, and on Cyprus. For further discussion, see
Muscarella, Phrygian Fibulae from Gordion, 59 f., Appendix C, 80f.; idem AJA 77 (1973), 236;
Jantzen, Samos VIII, 48 f. (Through a misadventure I reversed the numbers five and eight
when discussing the Phrygian and East Greek belt buckles published by Jantzen and men-
tioned in my review in AJA 77 (1973), 236; there are in fact five buckles with cast mouldings,
eight with added ones, not vice-versa as published. Thus there are, as listed above, apparently
eleven Phrygian objects on Samos beside the sherds. Fibula B473 (misprinted in the review
as 453), at least three belts (pl. 47), and two plaques (pl. 49), are probably East Greek. These
six objects plus the eight buckles with added mouldings are local, East Greek; which leaves
six objects listed by Jantzen as Phrygian as stili of unknown origin.)
Assyrian imports to the West include, from Sames, six statuettes (B1218, B165, B1594, B773,
B1214, B779), very probably another (B1217), four horse bits (pl. 61), for a total of ten or eleven
objects (AJA 77 [1973], 237). There is also a horse bit from Lindos (Herrmann, Pferdeschmuck,
22f., fig. 17), and a cylinder seal from Olympia, E. Unger in Max Ebert, Reallexikon der
Vorgeschichte 4/2 (Berlin, 1926), Taf. 162, c. There is also the stele of Sargon from Cyprus;
P. Demargne, The Birth of Greek Art (New York, 1964), 277; F. Dunbabin, Greeks, 41, refers to a
bronze plaque from Olympia as an object that may be Assyrian. To my mind this plaque is
North Syrian; Muscarella, Near Eastern Bronzes, 116 f., no. 2. On page 39 and note 5 Dunbabin
gives the impression that he considers the distinctly Assyrian seal from Olympia (above) to be
north Syrian, but this may be a misunderstanding on my part. Whether the Assyrian objects
recovered in the West were sent by Assyrian kings is a question worth considering; certainly
the Sargon stele on Cyprus would seem to have resulted from a royal command.
That Iranian objects and motifs passed over land and sea and that they
eventually touched Greek soil has been established. And we may claim with
some certainty that our historical knowledge of the West is broadened, for
we now know that there was a distinguishable Iranian spice in the unique
stew of Greek culture.
Addenda
After this paper went to press I came across J. Bouzek, Macedonian Bronzes:
Their Origins, Distribution and Relation to Other Cultural Groups of the
Early Iron Age, Pamtky archeologik 65 (1974), 278341. Bouzek states, as
I do in the text above, that the pomegranate shaped bell and the crea-
tures on a shank attached to a disc have an Iranian background (305, 323,
330); also that the Assyrian horsebits from Samos and Lindos are not Ira-
nian. However, contrary to the position taken by me, he claims that Samos
bronzes B1211 and B1130, and the maceheads/furniture attachments are Ira-
nian (323).
To Note 2:
An Achaemenid horsebit was excavated at Athens; H. Lechat, BCH 14 (1890),
385f., fig. 1.
Fig. 1. Open-work Pendant from Fortetsa, Crete.
Fig. 2. Spouted Bronze Vessel from Samos.

Fig. 3. Bronze Finial From Samos. Fig. 4. Bronze Mountain

Goat from Samos.
Fig. 5. Bronze Fig. 6. Bronze

Bell from Samos. Pendant from Samos.
Fig. 7. Bronze Goat from Samos.
Fig. 8. Bronze Horsebits from Delphi.

Fig. 9. Bronze Horsebit Attributed to Iran,

Purchased, Metropolitan Museum of Art, 56.42.2.
Fig. 10. Bronze Objects from Lindos, Rhodes.
Fig. 11. Bronze Seal Attributed to Iran, Private

Collection, Metropolitan Museum of Art, L61.77.
Fig. 12. Terracotta Vase, Louvre.

Fig. 13. Bronze Vase Attributed to Iran,

Fig. 14. Bronze Vase Attributed to Iran, Heeramaneck Collection.

Fig. 15. Bronze Spout from Aetos, Ithaca.
Fig. 16. Terracotta vessel from Hasanlu, Iran.

Fig. 17. Bronze Bell Fig. 18. Bronze Mace/Cosmetic

Attributed to Iran, Private Container from Hasanlu, Iran,
Collection, Metropolitan Metropolitan Museum of Art, 61.100.18.
Museum of Art, L59.41.4.
Fig. 19. Bronze Bells from Samos.
Fig. 20. Bronze Vessel Attributed to Iran,

Purchased, Metropolitan Museum of Art, 53.128.
Fig. 21. Double-Headed Animals from Athens and Olympia, Greece.
Fig. 22. Bronze Object Attributed to Iran,

chapter twenty-two
URARTIAN BELLS AND SAMOS*
Archaeological remains indicate that from the very beginning of the Greek
and Cretan Bronze Age, Aegean cities maintained contact with neighboring
lands in Egypt and the Near East. These contacts are indicated by imported
objects and materials excavated at sites on the Greek mainland and on var-
ious islands in the Aegean Sea, and, especially during the Late Bronze Age,
by Greek material found in Egypt and the East. After the collapse of the
Mycenaean civilization and the onslaught of the Dorian invasions, some-
time in the twelfth century bc, foreign imports are no longer encountered
on Greek sites, and there is no evidence that Greek goods continued to be
exported. These facts, coupled with the lack of monumental architecture
and the sparse remains from Greek sites, clearly indicate that Greece suf-
fered a cultural decline and that ties to the East had been severed. A change
occurs at least by the tenth century bc, if not slightly earlier, for a number
of sites in the Near East have yielded Greek Protogeometric pottery, a sure
indication that earlier contacts were being renewed. But during this period
the flow of goods apparently moved in only one direction, West to East, and
it was not until the late ninth century, at the earliest, that a reciprocal rela-
tionship began. It is at this time that the earliest near eastern material is
recovered in post-Mycenaean Greece, in particular at Athens. Aside from
these early findsa bronze decorated bowl, faience beads and plaques,
and gold earrings either imported or made under near eastern influence
the archaeological record is silent until approximately the mid-eighth cen-
tury bc. Beginning at this time, and continuing for some decades of the
seventh century, there is a quickening of imports to Greece, and it is dur-
ing this period that the greatest number of near eastern objects arrived.1
* This article originally appeared as Urartian Bells and Samos, Journal of the Ancient
Near East Society of Columbia University 10 (1978): 6166.

1 For summaries, discussions and bibliography of the imports from this period see O.W.
Muscarella, Near Eastern Bronzes in the West: The Question of Origin, in S. Doehringer,
D.G. Mitten, and A. Steinberg, eds., Art and Technology (Cambridge, Mass., 1970). 109f.;
H.V. Herrmann, Hellas (orientalischer Import im frhen Griechenland), Reallexikon der
Assyriologie (1975), 4:303311.
690 chapter twenty-two
It is one thing to recognize that a particular object recovered on Greek

soil is non-Greek, that it is an import from the Near East, but it is quite
another to determine accurately and to everyones satisfaction the specific
area or culture that produced the foreign objectalthough this would yield
valuable insight with respect to those near eastern centers that attracted
the Greeks and were involved in their spiritual and material development.
Indeed, from the very first discussions of imports in the nineteenth century,
scholarly opinion concerning whence they derived has been expressed,
and with the growth of our knowledge of near eastern material remains,
these opinions have of course been modified and refined. Nevertheless,
the difficulties involved in attributions are underscored by the fact that
consensus has not been reached concerning the specific place of origin of
every near eastern object and motif encountered in the West. In particular,
a great deal of the disagreement on origins has centered on studies of the
griffin protomes and siren cauldron attachments found in some quantity at
many Greek sites, as well as in Italy.
For many years almost every scholar who studied the siren attachments
has concluded that they were imported from Urartu and that the locally
made copies represent evidence for the penetration of Urartian art into
Greece;2 only a few earlier writers suggested Phoenicia or Assyria as the
source.3 Likewise many scholars also claimed an Urartian background for
the griffin protomes, seeing them either as imports or as examples of
2 There are approximately 79 examples known to date. From the Greek mainland there
are 38 oriental and 9 Greek examples, all excavated at sanctuaries, plus 3 Greek and 1 oriental
examples in museums that are attributed to Greece (a Greek example in Boston is the mate
of an example from Olympia); Delos and Rhodes each have one oriental type, and 6 are
from Italy, from tombs. Two aberrant examples were excavated at Olympia and 4 aberrant
examples come from Salamis, Cyprus. From the Near East there are 8 sirens from Gordion
and one example reported from Alishar, on the Araxes River. A total of 10 examples have
been attributed to the Van area but all derive from dealers or private owners; one example
has been attributed to Nimrod. Recently, the Munich art market had for sale a cauldron with
siren attachments in situ of a type hitherto not encountered. For relevant bibliography see
note 3, below.
3 Bibliography and discussions in detail concerning the origins of sirens and griffins are
to be found in Muscarella, Near Eastern Bronzes, 109 f., 116; idem, The Oriental Origin of
Siren Cauldron Attachments, Hesperia 31/4 (1962), 317329; H.-V. Herrmann, Die Kessel der
orientalisierenden Zeit (Berlin, 1966), 25 f., 50 f., 131 f.; idem, Urartu und Griechenland, JdI 81
(1966), 79141; idem, Hellas, 303 f., 306 f.; P. Amandry, Gnomon 41 (1969), 796f.; V. Karageorghis,
Excavations in the Nekropolis of Salamis III (Nicosia, 1973), 97f.; P. Amandry, LArt ourartien
et ses relations avec le monde grec, Proc. of the Xth Inter. Congress of Classical Arch. (Ankara,
1978), 5 f.
urartian bells and samos 691
Urartian influences.4 But far more important than the issue of the origin
of these groups of objects per se is the fact that most scholars who have
championed their Urartian background have taken the position that Urartu
played a significant rolethe major role some would arguein the near
eastern encounter with Greek culture in its formative stage.
A vigorous reaction to this position, the Urartu-These, began to develop
in the 1960s, and a growing number of scholars maintain that the sirens
and griffins, and still many more objectsbronze vessels, horse blinkers,
plaques, ivories, and assorted motifsderived from North Syria, not from
Urartu. Therefore, it follows that it was not Urartu but the North Syrian
cities that played the crucial orientalizing role in Greece.5 In addition, the
contributions of Phrygia, Assyria, and Iran have been better defined.6
This writer has long maintained that there are no recognizable Urartian
objects or motifs on Greek soil, nor in the West in general, and that Urartu
seems, to my mind, to have played no recognizable role at all in the trans-
mission of oriental art and motifs to the West.7 H.-V. Herrmann has referred
to the alleged presence of Urartian material in the West as an archologis-
che Legende, and he notes dass Urartu beim Export nach Griechenland
keine nennenswerte Rolle gespielt zu haben scheint.8
4 See note 3 for references. U. Jantzen, Griechische Greifenkessel (Berlin, 1955), reported
about 250 examples; G.M.A. Hanfmann, Gnomon 29 (1957), 241246, added over 60 more;
others continue to be excavated (e.g., AA 1978, 3, 396, fig. 15). Over 220 have been excavated
on Greek soil, 24 from Italy, 5 from France, and one from Spain. One example was excavated at
Susa, two are in the Ankara Museum, one in Izmir, and 2 gold examples have been attributed
to the site of Ziwiye, in Iran.
5 Especially Muscarella, The Oriental Origin, and Near Eastern Bronzes; Herrmann, Die
Kessel, and Urartu und Griechenland. See also R.S. Young, A Bronze Bowl in Philadelphia,
JNES 26 (1967), 150; J. Muhly, Homer and the Phoenicians, Berytus 19 (1970), 49; I.J. Winter,
Phoenician and North Syrian Ivory Carving , Iraq 38 (1976), 14, 16f., 21; idem; Carved
Ivory Furniture Panels from Nimrud , MMAJour. 11 (1976), 42; idem, AJA 80 (1976), 202;
Amandry, L Art ourartien, 5 f. A few scholars still support an Urartian origin for sirens and
griffins; S. Korti-Konti, Orientalizing Bronze Work in Greece, Athens Annals of Archaeology
4/2 (1971), 281 f.; S. Salvadori, An Urartian Bronze Strip in a Private Collection, East and West
26, 1/2 (1976), 106; C. Burney, The Ancient Near East (Ithaca, 1977), 183; A.M. Bisi, Elements
anatoliens dans les bronzes nouragiques de sardaigne, Proc. of the Xth Inter. Congress of
Classical Arch., (Ankara, 1978), 356 f.
6 For Iran, with bibliography, see O.W. Muscarella, The Archaeological Evidence for
Relations Between Iran and Greece in the First Millennium B.C., JANES 9 (1977), 3148; for
Phrygia and Assyria, see 32 and 47, n. 67; also Herrmann, Hellas, 306, 310f.
7 Near Eastern Bronzes, 122; see also The Oriental Origin, 325f.
8 Urartu und Griechenland, 80; Hellas, 396. Note also Amandry, LArt ourartien, 7: on
ne trouve dans le monde mditerranen aucun objet qui soit indiscutablement dorigine
ourartienne.
The Urartu-These surfaced in force again in 1972 with the publication by

U. Jantzen of the Egyptian and near eastern finds from the Greek sanctuary
of Hera on the island of Samos. Jantzen claimed that twelve of the many near
eastern bronze objects excavated over the years on Samos were Urartian.9
Almost all the reviewers of Jantzens publication resolutely rejected the
Urartian attribution of at least eleven of these objects.10 Indeed, only one
of the bronzes was accepted by some of the reviewers as being of Urartian
manufacture: a statuette of a horned male figure with outstretched hands
(B1217). Herrmann, Brker-Klhn, Hanfmann, and Kyrieleis11 all agreed that
B1217 is Urartian, and only this writer expressed reservations, claiming that
the attribution was not proven, that at best B1217 is Urartian?, and that it
could be Assyrian. Moreover, to Herrmann, not only is the Samian bronze
B1217 the only example of an Urartian object from Samos, it is the only
example of Urartian workmanship recovered to date in the Greek world.12
This position, with the reservations about B1217, is one that I have accepted
and maintained up to the present. But now new information allows us to
modify this view slightly.
Jantzen published eight bronze bells all of which were attributed by
him to the Caucasus.13 This view has been generally accepted, although
a few writers have called attention to the wide distribution of bells and
suggested that the Samian bells could have come from several areas.14 One
9 U. Jantzen, Samos VIII, gyptische und orientalische Bronzen aus dem Heraion von
Samos (Bonn, 1972), 80 f., pls. 7679.

10 O.W. Muscarella, AJA 77 (1973), 237; H.-V. Herrmann, Gnomon 47 (1957), 399f.; J. Brker-
Klhn, OLZ 70/2 (1975), 545; Amandry, L Art ourartien, 4, 6; G.M.A. Hanfmann, in Bi. Or. 30,
3/4 (1973), 199, alone accepted Jantzens attributions.
11 See note 10 and H. Kyrieleis, Orientalische Bronzen aus Samos, AA 1969, 2, 166f.,
J. Brker-Klhn, Verkannte neuassyrische Bronze-Statuetten, Bagd. Mitt. 6 (1973), 248f.

12 It is not quite clear just how certain Herrmann is about the Urartian origin of B1217,
in Urartu und Griechenland, 126 f. he said of it das berechtigte Anwartschaft darauf hat, als
urartisches Werk zu gelten ; and in Gnomon 47 (1975), 400, he states that it bleibt als
gesichertes urartisches Importstck ; but in Hellas, 306, he seems more hesitant, B1217
mag eine vereinzelte Ausnahme sein, i.e., may be the only Urartian object known in the
West. Amandry in L Art ourartien, 6, also expressed reservations about the Urartian origin of
this piece; however, he suggests that a bronze horse from Samos, B492, could be Urartian, a
position I do not share.
13 Samos VIII, 81 f., pls. 79, 80. A total of 20 bells were excavated on Samos, 12 of which are
considered to be Greek made.

14 Herrmann, Hellas, 311, and Gnomon 47 (1975), 400; Brker-Klhn, OLZ 70 (1975), 545;
Muscarella, AJA 77 (1973), 237, and JANES 9 (1977), 39 f., n. 32, for further references; H. Mbius,
Kaukasisches Glocken in Samos, Marburger Studien (Darmstadt, 1938), 156f. Even J. Bouzek
accepted the Samos bells as Caucasian, Macedonian Bronzes , Pamtky archeologik 65
of the bells from Samos (B474) stands out from the rest (Figure 1): it has
eight facetted sides with a double row of rectangular apertures separated
by a raised horizontal ridge and with another ridge at the base; the top
area consists of a heavy rosette moulding on which is a solid loop set on
a low platform. At the time the bell was published two similar examples
were available for study. One (Figure 2) is the octagonal bell with a single
row of rectangular apertures and raised horizontal ridges, two at the base,
one at the top, reported to have been found at Alishar on the Araxes River
together with two other bells and a bull and siren cauldron attachment.15
The uncontrolled find, made by Kurds in 1859 in a rock chamber (a tomb?),
precludes any objective information concerning the nature of the deposit,
or whether all the objects were indeed found together. Nevertheless, the
octagonal bell is inscribed with the name of the Urartian king Argishti, Ar-
gi-is-ti-i -ri-is-hi From the arsenal of Argishti (whether the first or second is
not certain and need not concern us here), and the bull attachment claimed
as part of the find is certainly of Urartian manufacture. Further, one of the
other bells, uninscribed (Figure 3), is also octagonal, although with only two
rectangular apertures, and it is clearly related in form to the Argishti bell.
But apparently because the third bell seems to be a typical Caucasian type
(Figure 4), and because the other bells are fenestered, a feature which is
considered to be Caucasian,16 and because Alishar is situated on the border
of Urartu, in Transcaucasia, scholars failed to identify the facetted bells as
Urartian. In any event, and for whatever reasons, the Alishar Argishti bell
was not cited by those who studied the Samos bells.
The second bell of concern to us was excavated at the Urartian site of
Karmir Blur and was published in 1955 (Figure 5).17 It is octagonal and has
a single row of rectangular apertures and raised horizontal ridges, two at
the base and one below the top; in fact, it seems to be a mate to the Alishar
example, except that it is uninscribed. This bell was also ignored by those
seeking parallels for the Samos bells, and B474, along with the other seven
examples published by Jantzen, was assigned to the Caucasus.18
(1974), 305, 309, 323, 333, 335; idem, Graeco-Macedonian Bronzes (Prague, 1973), 84, 87, 89,
fig. 25: no. 4 is B474 (infra), but it is an inaccurate drawing.
15 See P. Calmeyer, Glocke, in Reallexikon der Assyriologie (1969), 3: 429f., fig. 5; B.B.
Piotrovskii, Urartu: The Kingdom of Van and Its Art (London, 1967), 82f., figs. 58, 59.
16 Of course, the Caucasian apertures are typically triangular in shape, not rectangular:
see Mbius, Kaukasisches Glocken, 159 f., pl 68:4, 511.

17 B.B. Piotrovskii, Karmir Blur 3 (Leningrad, 1955), 46, fig. 35; idem, Karmir Blur (Lenin-
grad, n.d.), pls. 56, 58.

18 Muscarella, AJA 77 (1973), 237, mentioned Urartu along with other areas as possible
Recently, new information has appeared that now allows us to conclude

without reservation that there is a distinguishable Urartian bell type, a
type that includes the Alishar and Karmir Blur examples and the bell from
Samos, B474. In 1977 Oktay Belli of Istanbul University published an article
on Urartian horse equipment;19 he included three bells, all derived from
illicit digging in eastern Turkey and presently in the Van Museum. Two of
these bells, said to come from near Dizgin Kale, are inscribed, one with the
name Argishti, Ar-gi-is-ti-i -ri-is-hi (Figure 6), the other with that of Menua

(ca. 805786 bc), the father of Argishti I (Figure 7); the third, uninscribed,
is said to come from Patnos (Figure 8). All three are octagonal with a single
row of rectangular apertures and raised horizontal ridges, at the base and
top. And all have another characteristic featureone shared by the Alishar
and Karmir Blur bells (and apparently by B474), as well as by the next bells to
be discussednamely, that the suspension cross bar that held the clapper,
usually made of iron, was inserted through holes at the top below the loop
and its two ends project slightly.
Finally, and here the best comes last, at least two other bells available
for study indicate beyond doubt that Samos B474 is of Urartian origin. In
1974 the Ashmolean Museum acquired on the art market a bronze bell (Fig-
ure 9)20 that is octagonal and has a double row of rectangular apertures
separated by a raised ridge and another at the base, and a heavy rosette
moulding at the top surmounted by a solid loop set on a double ring plat-
form; the photograph does not show the holes for the suspension bar, but
they surely are there. The bell has the same Menua inscription as that on the
Van bell just mentioned. Although narrower in width, it obviously shares all
the features of B474 from Samos.
The second bell of interest here is one that was recently donated to
the Metropolitan Museum of Art by Nathaniel Spear, Jr. (Figures 10, 11).21
sources for the Samos bells, but neither the Alishar nor the Karmir Blur bells were cited.
P.R.S. Moorey, A Catalogue of the Ancient Persian Bronzes in tbe Ashmolean Museum (Oxford,
1971), 138, mentioned both Urartu and the Caucasus as possible sources for the Samos bells
but he could not have had B474 in mind as it had not yet been published.
19 Oktay Belli, Van Bolge Mzesindeki ivi Yazil Urartu Tun Eserleri, Anadolu 45
(19761977), 198 f., fig. 8, said to come from Patnos, fig. 9 has the Menua inscription; fig. 10
has the Argishti inscription.
20 P.R.S. Moorey, Ancient Iran (Ashmolean Museum, 1975), 25, pl. XIV; its height is 8.5cm.
Moorey claims the bell came from Western Iran, but this attribution comes from a dealer
and therefore has no archaeological value.
21 N. Spear, Jr., A Treasury of Archaeological Bells (New York, 1978), 111f., figs. 121, 122; height
8.7 cm.
This bell is also octagonal with a double row of apertures and ridges in
the appropriate places, and it too has a heavy rosette moulding at the top
surmounted by a solid loop set on a low platform; the iron clapper is now
missing but the iron suspension bar is extant (Figure 11). And it is inscribed
with the name of Argishti, in the very same manner as on the Alishar and
Van bells. The Metropolitan Museum bell is clearly of the same type as
the Ashmolean bell, but more significantly, in all features, except for the
inscription and its sizeit is 7 millimeters shorter and ca. 2 to 3 millimeters
widerit is a duplicate of B474 from Samos.
Given the readily identifiable characteristics of the Urartian bronze bells
under review here, it should be relatively easy to recognize other exam-
ples, or adaptations, in the Westif they existand, indeed, at least one
adaptation of an Urartian bell excavated on Greek soil exists. Spear, who
published the Metropolitan Museum bell, as well as those from Alishar,22
perceptively called attention to their resemblance in all essential details to
a small bronze bell from Pherae in Thessaly, Greece (Figure 12). This bell
has seven facets, a row of rectangular apertures, a raised ridge at the base, a
heavy undecorated moulding at the top surmounted by a solid loop set on a
low platform, and the ends of the suspension bar project from its holes.23 To
my mind the Pherae bell is a copy of an Urartian bell, rather than an import.
If so, it is at present the first certain example of a local Greek copy of an
Urartian work. Perhaps it might be rash to conclude that the Pherae bell is
directly dependent on Samos B474, but it is relevant in this context to note
that there were extensive contacts between Macedonia and Thessaly and
Samos in the eighth and seventh centuries bc, as we know from the impor-
tant research of J. Bouzek.24 Thus it is archaeologically acceptable to posit
that the Pherae bell was made as a result of contact with Samos, reinforcing
Bouzeks conclusions. At the same time, Bouzek believes that there is good
evidence for suggesting that there was contact between Macedonia and the
Caucasus,25 which means that we must equally allow for a more direct trans-
mission.
22 Treasury, 110 f., figs. 118120.

23 Treasury, 158 f., fig. 185; height 5.1 cm. Holly Pittman called my attention to the Pherae
bell.
24 Macedonian Bronzes, 305, 317 f., 335, figs. 36.
25 Ibid., 328 f., 335. Note that Bouzek does not mention the Pherae bell; on p. 309 he says
that Genuine Caucasian bells are only known from Samos, but related pieces come from
Pherai , In Graeco-Macedonian Bronzes he cited B474 from Samos as Caucasian. Guenter
Kopcke called my attention to Bouzeks discussion of Macedonian-Samos connections.
At this point in our discussion two questions arise: How does the presence
of an Urartian bell on Samos affect the thesis that there are no examples of
Urartian art in the West? How does the recognition that the Pherae bell is
adapted from an Urartian prototype affect the conclusion that Urartu played
no recognizable role in the formation of the Greek orientalizing taste and
style?
Concerning the first question, it is of course obvious that the position
must be modified to include B474; at least one Urartian object reached the
Westor to those who accept the statuette B1217 from Samos as Urartian,
there are now two objects. But one bell (and one putative statuette) does not
make influence, and I believe it would be irresponsible to refer to Urartian
cultural contacts with the West in any manner other than as the evidence
permits. At present the evidence is minimal. Reacting to Jantzens strong
statement with regard to the alleged quantity of Urartian art on Samos, Herr-
mann said that -die von J. aufgestellte Behauptung das sich in Samos das
urartische Kunsthandwerk besonders deutlich dokumentiere (80), ist auf
jeden Falle unhaltbar.26 I see no reason to modify this conclusion. Further,
the conclusion that no Urartian work of art or motif has yet been recog-
nized on mainland Greeze at any important sanctuaryAthens, Olympia,
Delphia, Perachora, Argos, Sparta, Ptoion, Tegea, Ithaca, etc.still obtains
and is still significant.
With regard to the second question, I believe that here too a conservative,
cautionary posture should be maintained. Because there exists on Greek soil
one bell that was manufactured with an Urartian model in mind does not in
any meaningful way indicate that Urartu can now be included among those
near eastern cultures, the products of which stimulated Greek artists and
artisans.27 Yet the Pherae bell with its Urartianizing features exists, albeit
in an area north of the major Greek centers, and as such it deserves to be
cited as an examplethe sole example I believeof a Greek object copied
from an Urartian model.
26 Gnomon 47 (1975), 400. Further, given the evidence presently available, I see no need
to alter previously expressed conclusions that neither Urartian nor Iranian goods travelled
across Anatolia to the West; Muscarella, The Archaeological Evidence, 45f., nn. 62, 63.
27 Amandry, L Art ourartien, 7, stated it neatly: Mme si quelques produits des ateliers de
la rgion de Van sont parvenus en terre grecque, on ne dcle, ni dans lart grec du 8e sicle,
ni dans le premier art grec orientalisantcelui des styles protocorinthien et protoattique
aucun trait qui soit ourartien.
I wish to express my sincere thanks to Oktay Belli, V. Lukonin, B.B. Piotrovskii, B. Philip-
paki, U. Jantzen, P.R.S. Moorey, and Nathaniel Spear, Jr., for their courteous and prompt coop-
eration and their permission to publish the bells from the Van Museum, Alishar, Karmir Blur,
Pherae, Samos, and the Ashmolean Museum.
Figure 1. Bell from Samos, B474. Figure 2. Bell said to come

from Alishar, Argisht;
Hermitage Museum.
Figure 3. Bell said to come from Figure 4. Bell said to come from
Alishar; Hermitage Museum. Alishar; Hermitage Museum.
Figure 5. Bell from Karmir Blur. Figure 6. Bell in Van

Museum, Argishti.
Figure 6b.
Figure 7. Bell in Van Museum, Menua.

Figure 8. Bell in Van Museum.
Figure 9. Bell in Ashmolean Museum, 1974.357,

Courtesy of Visitors of the Ashmolean Museum.
Figure 10. Metropolitan Museum of Art,

1977.186; gift of Nathaniel Spear, Jr. (four views).
Figure 11. MMA 1977.186, bottom view.
Figure 12. Bell from Pherae, Thessaly; Athens National Museum.

chapter twenty-three
KING MIDAS OF PHRYGIA AND THE GREEKS*
There are no internal historical sources for the history of the people the
Greeks called Phrygians. The earliest reference to the Phrygians in western
historical sources occurs in the text of Herodotus who lived in the 5th
century bc (I.14, VII.73). Herodotus (1.14) records that King Midas of Phrygia
dedicated his royal throne to the Delphic sanctuary and that he himself
saw it still on view.1 The date of Midas dedication is specifically situated by
Herodotus to a time before the dedications of King Gyges of Lydia (reigned
ca. 685650 bc).2 The implication of Herodotus statement is that Midas
reigned at a time before Gyges, and this chronological positioning almost
certainly establishes a pre-early 7th century date for Herodotus Midas. That
Herodotus King Midas of Phrygia lived in the 8th century bc and is the very
same person cited by the late 8th century Assyrian King Sargon II (between
the years 717 and 709bc) as King Mita of Mushki has long been accepted by
a majority of scholars, including the earliest to write about the Phrygians3
(G. Rawlinson: according to Bittel 1950: 76, but without references; Winckler
1901: 136, 283f., 287; Koerte 1904: 9f., 17, 18ff., 21; Kroll 1932: 1538 f.; Friedrich
1941: 863; Bittel 1942: 67f.; 1950: 76; 1970: 135; Akurgal 1955: 113, 120 ff.; Goetze
1957: 202; Dunbabin 1957; 64; Mellink 1965: 317 f.; Muscarella 1967: 59 f., 72,
n. 1; Herrmann 1975: 310; Roller 1983: 300).4
Herodotus firm chronological statement allows us to accept as an his-
torical facta hard factthat it was the 8th century bc, King Mita of the
* This chapter originally appeared as King Midas of Phrygia and the Greeks, in Anatolia
and the Ancient Near East: Studies in Honor of Tahsin zg, eds. K. Emre, M. Mellink,
B. Hrouda, and N. zg (Ankara: Trk Tarih Kurumu, 1989), 3344.
1 Koerte 1904: 21, suggests that the throne was not Midas but was one sent for the gods
use; see also Friedrich 1941: 888; Kroll 1932: 1539.

2 M. Cogan, H. Tadrnor, Gyges and Assurbanipal, Orientalia 46, 1977, 78., n. 25.
3 Snodgrass 1971: 350, claims we cannot be certain of the equation; Young 1981: 261, implies
the same lack of certainty.

4 Most of these scholars also accept the equation that Mushki equals Phrygians; see also
Muscarella 1988: 179 f. Postgate 1973 is ambiguous. He translates Mushki as Phrygian from
Sargons letter but on p. 24, n. 4, he claims that he does not imply that Mushku and Phrygia
were identical.
704 chapter twenty-three
Mushki-Phrygians cited by Sargon II who sentsurely from his capital city,

Gordiona valuable artifact, a throne, to Greece. Concomitantly, we have
important historical information that may legitimately be extrapolated con-
cerning Phrygian relations with the Greeks in the late 8th century, both with
the Greek east and specifically with the Greek mainland. This information
indicates (or at least suggests) an impulse that reflects something more than
casual knowledge of Greek cultural affairs. It precisely informs us that King
Midas had specific knowledge of the Delphic oracle, that he was aware of
its powers, and, most significantly, that he decided to honor the Greek god,
and perhaps make use of his oracular function. For although it may not be
claimed as an historical reality, it does not offend the evidence on its own
terms to suggests that behind Herodotus narrative lies an historical event:
Midas consulted the Delphic oracle with regard to a problem faced by the
Phrygians and the throne was presented as a dedication related to the ora-
cle sought.5 That is, the throne may not have been sent solely as a freewill
gift, but rather is to be viewed as a kind of gift exchange, the throne for the
oracle.
Whether the throne was presented as a free-will gift or as a dedication
related to an oracle, surely Phrygians themselves accompanied it to Delphi.
Such a valuable gift-dedication would presumably be delivered by personal
representatives of the donor; and if indeed an oracle was involved, the oracle
seekers delegates would presumably wish to be present to receive it and take
the message home. Phrygians travelled to Greece in the last decades or years
of the 8th century bc.
Aside from the personal comment that the throne was worth seeing,
Herodotus did not describe it or mention its material (pace DeVries 1981: 33,
and Prayon 1987: 193). As a result, however, of recent research and conserva-
tion of the wooden furniture from Tumulus P and MM at Gordion (accom-
plished by Elizabeth Simpson), we have an insight concerning the thrones
material and decoration. The serving screens and the elaborate three-legged
table (the Pagoda Table) from Tumulus MM are intricately manufactured
and formed of different colored woods, boxwood and walnut, and they are
decorated in geometric patterns with inlays of a different colored wood,
juniper (Simpson 1983; 1986). The Tumulus P serving stand has both juniper
and yew inlays; and a stool from the same burial is made of two woods, box-
wood and yew (information from E. Simpson).
5 The oracle sought could have been related to Midas problems with the Assyrians, or
even later with the Kimmerians.

king midas of phrygia and the greeks 705
Both in the quality of manufacture and in the skill and artistic nature
of their designs the Phrygian wooden furniture represents some of the
most aestheticworth-seeingartifacts known from antiquity, and they
objectively inform us of the woodworking capabilities of the 8th century
Phrygians. I suggest that it is highly probable that the throne Midas sent to
Delphi was made of woods of different color, and was probably elaborately
inlaid in geometric patterns (an idea indepentetly reached by Jeffery 1976:
148). The throne may have been made in the very same woodworkers shops
that made the Tumulus P and MM furniture.
Two late Greek sources record that King Midas of Phrygia married a Greek
princess. Aristotle (fr. 611,37, ed. V. Rose) calls her Hermodike and says that
she cut/struck the earliest coinage for Kyme.6 Pollux (Onamastikon IX.83)
names her Demodike, the daughter of King Agamemnon of Kyme, and he
notes that she was but one among several others who were alleged to have
been the first to strike coins.7 Both sources cite Kyme in Aeolis, on the west
coast of Asia Minor, as the princess home and Pollux specifically identifies
her father as being king there.
Given the late date (albeit derived from earlier sources) of the accounts,
the fact that coinage is mentioned, and that there were presumably 7th cen-
tury, as well as 6th century Phrygian kings named Midas (Herodotus I.35;
Koerte 1904: 25f.), it remains uncertain that the Midas cited is the Midas-
Mita of the 8th century bc, and not a later one (although it is clear that
the early Midas-Mita was in the minds of the authors). Scholars are divided
between those who assume that this Midas is the 8th century king (a major-
ity: Koerte 1904: 21, 23; Kroll 1932: 1539; Friedrich 1941: 388; Akurgal 1955: 125;
Huxley 1959: 94.f.; Barnett 1967: 14; Boardman 1973: 86; Mellink 1965: 317;
Sams 1979: 47; Roller 1983: 300; Diakonoff and Neroznak 1985: xiii; Prayon
1987: 118, 193), and those who cautiously leave the question open (Dunbabin
1957: 64; Lejeune 1969: 188; Young 1981: 253), Jeffery (1976: 238) assumes he
was a later, ca. 600bc, Midas.
6 They say that Hermodike, wife of the Phrygian king Midas, was very beautiful and
also that she was wise and a skilled craftswoman and struck the earliest coinage for Kyme.
(Translated by Donald F. McCabe.)
7 Another example of local pride is the dispute about coinage, whether the first one to
strike it was Pheidon of Argos, or Demodike of Kyme (who was wife of Midas the Phrygian
and daughter of King Agammemnon of Kyme), or Erichthonios and Lykos at Arthens, or
the Lydians (as Xenophanes says), or the Naxians (as Aglosthenes thought). (Translated by
Donald F. McCabe.)
The connection of Hermodike/Demodike with the incipience of coinage

at Kyme is an important issue of concern for establishing the chronology of
the Midas she married. Many scholars date the incipience to the late 7th or
early 6th century bc (Kagan 1982: 343ff.), which, if correct, would indicate
a gap of a century between Midas-Mita and the events mentioned in the
marriage narratives. The problem of chronology, however, while relevant
and pertinent, is not insoluble. For, in the present state of our knowledge,
and as a result of recent research, it seems that the precise time when
coinage appeared in Asia Minor is by no means a settled archaeological
issue. D. Kagan (1982), using in part the work of L. Weidauer, has reviewed
the chronology of the Basis deposit at Ephesusa crucial locus for the
dating of early coinageand has demonstrated (what P. Jacobsthal, PHS
1951: 85ff., had previously concluded) that the Basis material contains 7th
century bc material, material earlier than the assumed ca. 600 bc date for its
final deposition. Kagan (1982: 347ff., 355f., 359) dates the deposition to ca.
the mid-7th century bc8 and, with Weidauer, he vigorously concludes that
the coins therein, and thus the beginning of coinage, could date close to the
beginning of the 7th century bc.9
Whether this early dating of the final deposition of the Ephesian Basis
will be supported by other scholars remains to be seen: Akurgal (1983) for
example, has already argued contra Weidauer that the earliest Ionian coins
do not pre-date the last quarter of the 7th century bc.10 Nevertheless, Kagan
(and Weidauer) has forced us to consider that the earliest coins in the
Basis need not be so late as 600bc. The 4th century bc historians believed
that coinage existed in the 7th century but, as the passage of Pollux makes
clear, they were unsure where in place, and under whose guidance, coinage
first originated. Coins inscribed with the name Phanes, assigned either
to Ephesus (Jeffery 1976: 222) or Halicarnassus (Balmuth 1971: 6), may be
ca. 625 bc in date, and because the name is not that of a king it raises issues
regarding moneyers other that heads of state. Moreover, if by coinage one
signifies fabricated pieces of controlled weight supplied with a name or
logo that guarantees weight, then the silver ingots inscribed with the name
of the 8th century king Barrekub from Zincirli could strictly be considered
8 Kagan on p. 359 cites Weidauer as dating the deposition to 626bc at the latest, but she
too assigns an earlier date for the coinage.

9 See the rebuttal to Kagan by J.H. Kroll and N.M. Waggoner in AJA 88, 3, 1984: 325ff., but
they speak only to the issue of the dating of Greek coins.

10 Kroll and Waggoner (see note 9) refer to a congress in London on the chronology of the
Basis of Ephesus but it has not yet been published.

coinage (Balmuth 1971: 5ff., pls. 2, 3). The thrust of all this is to demonstrate
that scholars are not in a strong position when they categorically deny the
existence of some form of coinage in Asia Minor in the 7th century bc, earlier
than 625 bc (as Akurgal 1983).
It follows from the above comments that it is not a solecism if one accepts
as viable the conclusion that Midas-Mira married a Greek princess in the
late 8th century bc (or a few years into the 7th century)a princess who
may have played a role in the introduction of coinage to Kyme at some
time (still to be determined) after the death of her husband (ca. 696bc).
The Kymean princess could have been a young woman when she married
Midas-Mita and a middle-aged or older widow when she involved herself
in the economic affairs of her home city. If we can trust some elements
in the pseudo-Herodotus Life of Homer11which offers further information
about Midas marriageMidas inlaws, presumably King Agamemnon and
his queen, survived his death (that is, if we can trust this part of the story,
if being sceptical of the rest), for they are supposed to have commissioned
an epitaph for his tomb.12 If they survived Midas death, so could their
daughterwho presumably would have left Gordion at the threat of or as
a result of the Kimmerian invasion. Another possibility to be considered,
but one that cannot be forced, is that the coinage connection is an error or
a later misunderstanding (Huxley 1959: 94), joined somehow to an original
account of the marriage.
A second issue of concern for the marriage is the chronology of Agamem-
non. Koerte (1903: 19) and Wade-Gery (1952: 7) date him to ca. 700 bc, but
their chronology seems based on the acceptance of his being the father-in-
law of the 8th century King Midas (see also Roller 1983: 300; cf. Jeffery 1976:
238). If it can be firmly established that kingship at Kyme did not continue
into the 7th century bc (v. Bchner, PW, supp. 1, 1903, Kyme, 2475), we might
have in the reference to a king in Kyme a good indication that he was a con-
temporary of Midas, and could have been his father-in-law.
The problem may be approached from another direction, one that I
suggest supports the claim for a late 8th century marriage. For this evidence
we return to the throne dedication. An archaeological interpretation of the
11 Vita Homeri Herodotea in Homeri Opera, ed. T.W. Allen, vol. V, 198, lines 130140, for the
epitaph commission (Donald McCabe located this reference for me); Jeffery 1976: 198. Plato,
Phaedrus, 264D, gives a shorter version; see Koerte 1904: 22, and Wade-Gery 1952: 65, n. 21.
12 Koerte 1904: 22, considered the epitaph to be a fiction. There is no discussion in the
ancient literature concerning where the alleged tomb was located.

dedication and the dynamics informing it not only reinforces the tradition
of the marriage, it also yields information about the nature and the degree
of Phrygian-Greek relations.
Although it is not an impossible view to uphold that the Phrygians
learned of the Delphic oracle from the Cilician or north Syrian coastal cities
(e.g. the ubiquitously cited al Mina), there are no compelling indications
for such an assumption, and there is no historical or archaeological evi-
dence for direct Phrygian contacts there.13 Phrygia indeed had political and
cultural contacts with inland cities of north Syria, especially Carchemish,
with whose king, Pisiris, Midas made a short-lived treaty in 717 bc (Lucken-
bill 1927: para. 8). Undoubted north Syrian artifacts and motifs have been
recovered at Gordion, and north Syrian motifs influenced Phrygian potters,
collectively attesting to this relationship (Muscarella 1967: 67 f.; Sams 1974:
181ff., 191; 1978: 233ff.; 1979: 46; Prayon 1987: 183 ff.). Except for a sherd at Tell
Halaf, no certain Phrygian material has been recovered in north Syria or the
Levant.14 And the Phrygians never penetrated to the Cilician coast, because
of Assyrian interference (Luckenbill 1927: para. 18, 42; Postgate 1973: 28, 30,
32). There is no evidence that in the 8th century bc Phrygia maintained trade
routes to the Cilician or Syrian coasts.
From an archaeological and geographical perspective it makes more
sense to recognize that the Phrygians learned of the oracleand other
Greek mattersfrom the west coast Greek cities, a region that was closer
than the sourthern ports, and where there were no geographical or mani-
fest political barriers to contact (see Koerte 1904: 21 ff.; Friedrich 1941: 888).
Given these realities, joined with the archaeological record (infra), it seems
certain that Midas throne must have travelled (by cart!) from Gordion via
an overland route west to the Greek coastand thence by ship to Greece.15
Further, a considerable number of Phrygian artifacts were recovered in the
Heraion sanctuary on Samos (infra), and Boehmer (1973: 166) has cogently
suggested that some of this material represents 8th century dedications of
13 A XII, 14 sub-type fibula was excavated at al Mina, but it is clearly post-8th century bc
in date: Boehmer 1972: 63. Birmingham 1961: 189, incorrectly cites a XII, 13 fibula from Zincirli.
14 Akurgal 1966: 118, figs. 9092, claims that bronze vessels from Tell Halaf are Phrygian, but
none of the examples cited has a Phrygian parallel, and handles at right angles to the spout
occur outside of Phrygia. In 1967: 67, I cited a painted jug from Carchemish as Phrygian: Sams
1974: 192; 1978: 235, seems to reject the attribution. I also here reject my 1967: 67, comments
regarding right angled handles to spouts that occur in north Syria as definitely Phrygian.
15 Pace Winckler 1901: 284 f. (see also Kroll 1932: 1534) there is no evidence that Phrygia
was a sea power. The ship that transported the throne was probably Greek.
Midas to Hera. These dedications (whether by Midas or other Phrygians)

must have reached Samos from the neighboring coast en route from Phrygia,
and not from the Levant.
If Herodotus account of the Phrygian dedication to Delphi is fully com-
prehended, it not only indicates knowledge of Greek affairs, and here distant
Greek affairs, it also implies, one may even claim it presupposes, friendly
relations between the Phrygians and more than one Greek state. And this
knowledge and friendship was surely easier to acquire to Phrygias west than
to its south. Such an interpretation fits well with the suggestion that Midas
would have sought to maintain political alliances to his west: which the
marriage tradition neatly illustrates (see Koerte 1904: 21 ff.; Kroll 1932: 1539;
Akurgal 1955: 125; Goetze 1957: 202; Muscarella 1967: 63), for marriage is a typ-
ical Near Eastern custom employed to strengthen a treaty between states.
Thus, rather than creating a circular argument, the marriage narratives and
the throne dedication support each other to form a coherent historical sce-
nario about Phrygian-Greek alliances and friendship.16
There is solid archaeological evidence at Gordion and elsewhere that
incontestably documents some form of contact between the Greek coast
and Phrygia in the time of Midas reign, evidence that surprisingly has been
ignored. Eight leech-type fibulae of the type used in the Greek islands and
on the west coast of Asia Minor were excavated in the destruction level at
Gordion and in two 8th century tumuli, G and Y (Muscarella 1967: 62, 82;
Caner 1983: 41f., pl. 74A). These fibulae are manifest East Greek imports
at Gordion, imports that surely reached there from the west coast, not
from Cilicia or the Levant.17 Additional relevant evidence for 8th century
Phrygian contact with the west consists of two East Greek copies of Phrygian
fibulae that were excavated at Pithecusae on the Italian island of Ischia,
dated almost certainly to the late 8th century bc. That East Greeks knew of
Phrygian fibulae and copied them at this time (Muscarella 1967: 62; Boehmer
1972: 60, n. 424) is a further indication of an overland route between the
two areas. An overland route to the west is an archaeological reality, one
recognized years ago by Koerte (1904: 21ff.) and Kroll (1932: 1539), even before
the Gordion data were available.18
16 We know nothing of the dynamics of the relationship: who initiated it; how many Greek
cities were involved; how much physical contact occurred. Nor do we know if the marriage
preceded the Delphic dedication.
17 These fibulae are not mentioned by Sams 1979: 47; Young 1963, Roller 1983: 301 or Prayon
1987: 193 ff.

18 See also Muscarella 1967: 63; Lejeune 1969: 188. Most of the literature that supports an
A small number of Greek sherds, but representing five or six vessels, of 8th
century bc date have been recovered at Gordion. Although they all derive
from later levels (Sams 1979: 47) it is highly probable, as Sams states, that
they were displaced from the destruction level and in fact reached Gordion
during the reign of Midas. Sams prefers to see the vessels as coming from
the south, from Cilicia or the Levant, not the westwhich he admits is
the most obvious channel; but his preferred direction warrants review:
more probably, most likely, the vessels travelled overland from the west, the
obvious channelalong the same route that carried the fibulae and the
throne, and more material still to be discussed.19
Over this same route from the Greek east was most probably transmitted
the alphabet, from Greek minds to Phrygians (because the Greeks called
their alphabet Phoenician, not Phrygian), although the opposite direction
has been considered (Barnett 1967: 20; Young 1963: 363, no. 54; 1969: 264,
294: but cf. p. 296). Whence it reached Phrygia is of course in great part a
philological matter, and the archaeologist takes risks in offering opinions.
Young (1969: 253ff., 264 f. 294) vigorously argued for Cilicia or the north
Syrian coast and rejected the west coast of Asia Minor for the learning area
(see also Snodgrass 1971: 352; Sams 1974: 192). Diakonoff and Neroznak (1985:
3) on the other hand argued for the opposite view; they rejected the Levant.
From an archaeological perspective, Diakonoff and Neroznaks thesis best
fits into the background, and it does not contradict the philological data.
The transmission of the alphabet from Greece to Phrygia, the fact that
Phrygians chose this alphabet in which to communicate (perhaps because
of a common ancestral background?), rather than the Assyrian cuneiform
(adopted by the Urartians), or the Luwian hieroglyphs (employed in Tabal
and north Syria), remains the most significant cultural event resulting from
the relationship between the two people: and the Phrygian choice manifests
the closeness of that relationship.
overland route across Anatolia has been concerned with explaining how alleged Urartian
material allegedly reached the west, e.g. Birmingham 1961, and others; see Muscarella 1977:
45 f.
19 I must citebut wish to reserve judgment onthe conclusions of A. Snodgrass and
Prayon (1987: 163, nn. 677680, pl. 33d) that a horseman depicted on an ivory plaque from
pre-destructions Gordion wears a Greek helmet, specifically the closed Corinthian form (to
Snodgrass he is a Greek, to Prayon he is a Phrygian wearing a Greek (style) helmet). If indeed
the helmet can be explained only in Greek terms (that the artists rendition and intention
is not misunderstood by us), then the plaque would be an important piece of evidence
for a Greek cultural presence at Gordion. At present, I do not see the certainty of a Greek
connection, hence its parenthetic inclusion here.
We turn our attention now to a review of Phrygian artifacts archaeolog-

ically encountered in the west, in Greek controlled areas. Four categories
of artifacts, all constructed of bronze, comprise the Phrygian corpus: fibu-
lae, belts and belt handles, omphalos bowls, and cauldrons with bull head
attachments. All the Phrygian material derives from sanctuaries, that is from
public, not private loci. They are manifestly votive offerings and are found
alongside the offerings of Greeks and with material brought from identifi-
able, and perhaps from other, still to be identified, areas in the Near East.
With each of the Phrygian categories certain problems confront us: their
chronological range within Phrygia itself, the precise chronology of their
occurrence in the Greek sites, and obtaining a correct stylistic determina-
tion concerning whether the artifacts are actual exports or Greek imitations
of these exports.20
Fibulae are the best known and the most published examples of Phrygian
artifacts excavated in west. Following the classification of Blinkenberg (1926:
204ff., Type XII), Muscarella (1967: 12ff.), and Boehmer (1972: 46 ff.), the
forms of Phrygian fibulae found in the west consist primarily of sub-types
XII, 5, XII, 7, XII, 9 ( and of Boehmer), XII, 13, and XII, 14.21 (For a listing
of the find spots with bibliography of all Phrygian fibulae found in the
west, including post-8th century examples, see Muscarella 1967: chapter 2,
appendix C, and Boehmer 1972: 46ff.)
XII, 5 fibulae occur in Phrygia only at Gordion, in an 8th century tumulus
(KIII), and in the destruction level on the city mound (ca. 696 bc); a stray
was found in the latest city mound level. XII, 7 fibulae occur in the late
8th century Tumulus MM at Gordion (and probably also in the 8th century
tumuli KIII and KIV: Muscarella 1982: 198); a single stray occurs in a late level.
At Ankara a double-pin example occurs in an 8th century tumulus (Caner
1983: 14). Bogazky has double-pin examples dated to the 8th century and a
single-pin ones dated to the 7th (Boehmer 1972: 52 ff., nos. 7982A; Neve 1972:
183f., fig. 11). Thus at Gordion both sub-types are 8th century forms but XII, 7
are attested later at Bogazky (Muscarella 1967: 16 f.; Boehmer 1972: 52; Caner
1983: 68, 105). In the Greek sphere, 8th century XII, 7 forms (with sharply cut
20 While the imitations are culturally significant in terms of Greek adaptations of Phrygian
material, they are not manifestly dated to the 8th century bc; the Greek artisans may of course
have used 8th century Phrygian models.
21 Strangely, not a single example of XII, 7A fibulae, a common sub-type, occurs outside of
Phrygia. Note, however, that a broken fibula arc from Argos (Waldstein 1905, pl. 87, no. 903)
could be a XII, 7A form as the arc seems to be a crescent and the end tapers. Neither Boehmer,
Caner, nor the writer have cited the fragment.
mouldings) come from Lindos on Rhodes, and Samos (with double pins),
while XII, 5 fibulae occur at a number of island and mainland sites.
XII, 9 fibulae, a sub-type of XII, 7, embellished with studs on the arc,
occur in three 8th century tumuli at Gordion, MM, KIII, and KIV, but also
in Tumulus S1, which may be post-destruction in date; they are also found
in the destruction level and in almost all post-destruction levels on the
city mound. At Bogazky they occur in levels dated to the 8th and the 7th
centuries bc (Boehmer 1972: 56). In Ankara some were recovered from an
8th century tumulus (Bulu 1979: pl. 9). XII, 9 fibulae were also dedicated on
Greek island and mainland sites. An example from Marino in Latium may
have reached Italy by indirect trade, from a Greek source; it was deposited
in a 7th century grave (Muscarella 1967: 19, no. 26; Boehmer 1972: 57, n. 394).
While the majority are dated in Phrygia to the 8th century bc, enough
examples are attested there from later levels to indicate that the sub-type
continued to be manufactured in the 7th century, and apparently later.
R.M. Boehmer (1972: 50f., n. 323) has thoroughly reviewed the chronology
of the find spots at the Greek sites where Phrygian fibulae have been recov-
ered. His conclusion is that none of these sites has a clear-cut context that
allows one to argue with certainty that a Phrygian fibula reached there in the
8th century bc rather than in the 7th: viz. Sparta, Lindos, Argive Heraeum,
Perachora, Olympia, Samos; and that at other sites, viz. Chios, Aetos (Ithaca),
Lesbos, they are 7th century arrivals.22 At the same time, however, on the
basis of formal similarities to those from Phrygian 8th century contexts, he
accepts an 8th century date for a double-pin XII, 7 fibula from Samos, two
XII, 7 examples from Lindos, and the same date for XII, 9 fibulae from Lin-
dos and the Argive Heraeum (Boehmer 1972: 53, 54, 57, nn. 388, 392, 396; 1973:
166). I agree, and I would further suggest that a XII, 9 example from Olympia
and one from Paros, and perhaps some (if not all) of the XII, 5 fibulae from
the Argive Heraeum, Sparta, and perhaps Lindos could be 8th century bc
arrivals there. All the examples cited here could very well be curated 8th
century dedications placed with later material in the sanctuary deposits, a
view accepted, in part at least, by Boehmer.
Sub-types XII, 13 and XII, 14 fibulae are among the most common Phry-
gian examples excavated both within Phrygia and in the Greek sphere.
They occur in quantitiy at Gordion in 8th century tumuli and in all the
22 Boehmer (1972: 57, n. 397) considers some examples of his XII, 9 , found at some of
these sites, to be local copies, and therefore post-8th century bc. Of course, we do not know
whether this is true or not.
post-destruction city mound levels (Muscarella 1967: 21 ff., 24 f.). At Bogazky

a few examples of these sub-types occur in late 8th century bc contexts
(Boehmer 1972: 59ff., 62ff., Neve 1972: 183f., fig. 10). Their chronology has a
longer range than that of any other sub-type, a fact that complicates dating
the many examples found abroad, all the more so as many were copied in
the west.23 Boehmer dates practically all examples of those found in the west
to the 7th century bc or later; he allows a XII, 13 example from Lindos to be
late 8th7th century, a XII, 14, and a XII, 14A (a variant of XII, 13) example
from the Argive Heraeum to be possibly late 8th century in date, but here
too he prefers a later date (Boehmer 1972: 60, 63, no. 496, 65, no. 504).24
It is difficult to resolve the issue of the western chronology of XII, 13
and XII, 14 fibulae. Subjective interpretations regarding the comparison
of western forms with known 8th century examples in the homeland are
involved.25 Yet, we know from the Pithecusae fibulae (supra) that XII, 13
fibulae were copied by East Greeks in the 8th century bc; and there are
at least two (unpublished) XII, 13 fibulae in the Olympia Museum that
have double pins, which feature in Phrygia seems to be confined to the 8th
century (Boehmer 1972: 53; 1973: 166). This evidence enables one to suggest
with some conviction that a number of XII, 13 and XII, 14 fibulae reached the
Greek sanctuaries in late 8th century bc.
To summarize, it may be stated with some degree of certainty that in the
time of Midas reign Phrygian fibulae were dedicated at a number of Greek
sanctuaries, at least Lindos, Samos, Olympia, the Argive Heraeum, and per-
haps Paros and Sparta. Whether these fibulae were dedicated together with
Phrygian clothing, to which they were originally associated, remains an
interesting possibility (Boehmer 1973: 149ff., 166 ff., 172; 1983: 75, 78; Herr-
mann 1975: 303, 310; Prayon 1987: 193, n. 844). What remains secure, however,
is that fibulae per se were votive objects (Muscarella 1967: 52, 56, n. 16; Caner
1983: 67, 201f.) and they could have been dedicated at the sanctuaries inde-
pendently of clothing.
Phrygian bronze belts, elaborately decorated with geometric motifs, and
with a buckle form similar to fibula, occur at Gordion in the 8th century
Tumulus P (Muscarella 1967: 26, 51, pl. XV, 77, 78; Young 1981: figs. 911, pls.
23 Pn copies of Phrygian fibulae see Muscarella 1967: 37ff.; Boehmer 1972: 50, 58, 60, 63,
1973: 149 f.
24 Boehmer (1973: 149 f.) publishes a terracotta fragment from Argos with a XII, 14 fibula
in relief; its date is unclear.

25 For example, I could see Waldstein 1905, no. 983 as being early, whereas Boehmer (1972:
63, n. 496) sees it as 7th century.

11, 12), at Bogazky (Boehmer 1972: 72f., no. 180, dated to the 7th century,
cf. Neve 1972: 183, fig. 11, dated to the 8th century), and at Ankara (zg
and Akok 1947: figs. 2326; Bulu 1979: pl. 14: 8th century). The Tumulus P
examples have handles with a plain arc, except for the terminal mouldings;
the example from Bogazky has a central moulding; Ankara lacks a handle.
Other, later, examples from Gordion have multiple mouldings on the arc
(Muscarella 1967: pl. XV, 77, 78).26 A number of belts, or buckles with missing
belts, have been excavated in the west at Chios, Ephesus, Bayrakli, and
Samos (Boardman 1961: 179ff.; Jantzen 1972: pls. 4547). Except for some
handles and a belt fragment from Samos (Jantzen 1972: B593, 606, 613, 616,
1691, 1289),27 all the others seem to be East Greek copies of Phrygian belts.28
And it must be admitted that all the buckles from Samos could be post-
8th century bc in date. They all have multiple mouldings on the arc, a
late feature at Gordion (Muscarella 1967: 26), although not apparently at
Bogazky (see note 26 here).
Typical Phrygian omphalos bowls and bowls with bolster attachments
and ring handles joined to horizontal rim bands with vertical spools, known
from Gordion and Ankara, have been excavated in the west, in some cases
at the same sites that preserved Phrygian fibulae. Although in the past there
has been confusion regarding which bowls are actually 8th century Phrygian
imports and which local adaptions, the full publication of three 8th century
tumuli at Gordion (Young 1981) and the work of A. Knudsen (1964) has made
it easier to resolve these problems.
At least four Greek sanctuaries preserve examples of Phrygian omphaloi
and bowls with ring handles and bolsters: Samos, the Argive Heraeum,
and Perachora. From the Heraion on Samos derive the remains of two
ring-handled bowls (Jantzen 1972: 54, pl. 50, B494, B413), accepted as 8th
century bc by Jantzen, Knudsen (1964: 63, n. 10), Muscarella (1973: 236;
1977: 47, n. 67), and Herrmann (1975: 310). The remains of a horizontal rim
band and spool that derive from Olympia (Frtwangler 1890: 136, no. 852)
have been accepted equally as 8th century Phrygian by Knudsen (1964:
64, n. 10) and Herrmann (1975: 310); so have horizontal rims with spools
26 But compare a XII, 14A fibula, with multiple mouldings on the arc, from a grave dated
to the 8th century from Bogazky: Boehmer 1972: 64f., fig. 28.
27 Compare Samos B606 and B1691 to Boehmer 1972: no. 180, and fig. 28; and B613 and B616
to Muscarella 1967: pl. XV, 77, 78. Note the moulding formation and the ridged terminals. For
the Samos belt compare zg and Akok 1974, fig. 25, and Neve 1972: fig. 11.
28 Boardman 1961, 185, has noted the different form of the catch area: holes on the East
Greek Belts, loops on the Phrygian belts.

representing three bowls from the Argive Heraeum (Young 1963: 362; Herr-
mann 1975: 310).29 Finally, two plain omphaloi from Perachora (Payne 1940:
pl. 55, nos. 1 and 4) are most probably Phrygian imports (v. Dunbabin 1957:
152; Birmingham 1961: 190, n. 4; Muscarella 1967: 60, 72, n. 4: cf. these bowls
to Young 1981: pl. 72, MM 133, MM 137, MM 139, MM 141, MM 145, and pl. 90,
W 11, W 13, W 14).
Note also that bowls with banded rims and bolsters have been excavated
on Cyprus and at Magnesia in Asia Minor (Akurgal 1955: 81, 82, pl. 60a;
Knudsen 1964: 63).
In addition to the above, petalled decorated omphalos bowls have been
excavated at Olympia, Perachora, and the Argive Heraeum (Frtwangler
1890: pl. 52, no. 880; Payne 1940: pl. 52; Waldstein 1905: pl. 114, nos. 1975, 1976),
and rim bands or just spools derive from Samos (Jantzen 1972: pl. 50, B1397,
B1633), Lindos (Blinkenberg 1931: pl. 30, nos. 718, 719), and Delphi (Perdrizet
1908: 78, fig. 270). All of these bowls seem to be Greek imitations of Phrygian
examples, as noted by Knudsen (1964: 63f., no. 10), Muscarella (1973: 236),
Mellink (1981: 236, n. 82): but compare Akurgal (1955: 82), Birmingham (1961:
190), Muscarella (1967: 60; 1977: 47, n. 67), who cited some of these examples
as actual Phrygian imports.30
The bona fide Phrygian bowls cited above all match 8th century exam-
ples from Gordion, and it seems difficult to reject a conclusion that they
reached Greece during that time. At three of the sites that preserved Pry-
gian bowls, Samos, Olympia, and the Argive Heraeum, Phrygian fibulae were
also recovered. However (as Kenneth Sams reminds me), banded rim and
omphalos bowls occur at Gordion in post-destruction 7th century bc tumuli,
and to be archaeologically correct we cannot exclude the possibility that
some of the bowls reached Greece after Midas reign. So little is known of
post-destruction 7th century Phrygian history that it is not certain whether
Phrygia continued to maintain strong and direct relations with Greece. But
inasmuch as we are aware of 8th century contacts, at least it may be argued
that some of the bowls could have reached Greece at that time. No stronger
statement is warranted.
We now turn to the bronze cauldrons with bull head attachments, two
to a vessel. At Gordion two bull cauldrons derive from the 8th century
Tumulus W (W 1, W 2), three from Tumulus MM (MM 1, MM 12, MM 13),
29 For Gordion parallels to these spools and rims see Young 1981: pls. 60, 66: MM 60, MM 64.
30 Note that no bona fide examples of Phrygian bowls with petalled decoration have
been recognized in the west: but if local copies exist perhaps Phrygian imports were used
as models.
a single corroded example from Megaron 4, and a pair from a cauldron was
excavated in the Terrace Building in the destruction level of the city mound,
i.e. from the early years of the 7th century bc (Young 1981: 102, 112, 199 f.,
figs. 67, 68, 117, pls. 50, 59, 87, 88, 95B).31
The Gordion cauldrons have been attributed either to Phrygia or to north
Syria, the latter as imports. The two cauldrons from Tumulus W were con-
sidered to be local product by Young (1960: 231; 1981: 219), but north Syrian
by Herrmann (1966: 128; 1980: 576; 1984: 23), Kyrieleis (1977: 76), DeVries
(1981: 221), and Prayon (1987: 122ff.). The three Tumulus MM cauldrons were
accepted as Phrygian by Young (1958: 151), Akurgal (1961: 202), van Loon
(1966: 105, n. 119), DeVries (1981: 222), Prayon (1988: 124 ff.), and Muscarella
(1987: 185). Herrmann (1966: 84, n. 27, 122, 128, 129; 1980: 576 f.; 1984: 23)32
accepted MM 12 and MM 13 (he actually knew only one example at that
time) as Phrygian, but he considered MM 1 to be north Syrian; Kyrieleis
(1977: 76) apparently only considered MM 11 and MM 13 to be Phrygian; and
the Terrace Building pair were accepted as Phrygian by Kyrieleis, Herrmann,
DeVries, Prayon, and Muscarella. The Megaron 4 example may be North Syr-
ian (DeVries 1981: 221; Prayon 1987: 124).
At least MM 12 and MM 13, the Terrace Building pair, and possibly MM 1
may be accepted as probable Phrygian products, the characteristics of
which, especially the T-shaped clamp attachment, have been given by
Kyrieleis, DeVries, and Prayon.33
Herrmann (1966: 129; 1975: 310) considered it possible that bull attach-
ments from Argos (Waldstein 1905: pl. 75, no. 25), from Athens (de Ridder
1896: 198, fig. 164), one from Samos (Jantzen 1972: pl. 77, B1266), and three
examples from Olympia (Herrmann 1966: pl. 43, A25, A26, pl. 51, A33) were
also Phrygian exports. But except for the Samos example, which is accepted
as possibly Phrygian by Kyrieleis (1977: 87), DeVries (1981: 222), Prayon (1987:
125f., 211), and Muscarella (1988: 185), all the others are surely not: the Argos
example could be Near Eastern or Greek, and the Athens and Olympia exam-
ples are surely Greek (Muscarella 1968: 14, n. 20; Kyrieleis 1977: 87).34
31 Birmingham 1961: 191, refers to bull cauldrons from Ankarabut in fact she was exam-
ining MM 12 from Gordion. In Muscarella 1988: 185, I mention two examples from the METU
excavations at Ankara, but there is actually one example and of a type other than those dis-
cussed here: Bulu 1979: pl. 13, fig. 6.
32 Herrmann makes his determination here with caution; see also Muscarella 1968: 13f.
33 For other cauldrons with the T-shaped clamp see Young 1981: pl. 58; Bulu 1979: pl. 13,
figs. 4, 5. Kyrieleis and DeVries have made it easier to distinguish Phrygian from North Syrian
bull attachments, a problem I raised in my 1968 paper.
34 Note that in 1980: 577, Herrmann called the Argive example north Syrian; and he did not
Another cauldron with bull head attachments has been introduced into
the literature as being Phrygian, the well-known example in Copenhagen,
which although without provenience is usually assumed to have come from
Cumae in Italy (see Akurgal 1961: figs. 33, 34). While most scholars convinc-
ingly attribute the cauldron to a north Syrian workshop (Herrmann 1966:
123, 128, 129, n. 45; 1980: 576; Kyrieleis 1977: 74, 76; DeVries 1981: 221), both
Young (1960: 231) and Birmingham (1961: 191) have called it Phrygian. It is
surely not Phrygian and could very well be north Syrian (cf. Muscarella 1968:
14f.). To summarize, there is but one bull attachmentwhich equals one
cauldronfrom the west that seems to be a Phrygian export, the example
from Samos.
A review of Phrygian-Greek cultural relationship and contact must also
bring into the discussion a miscellaneous group of artifacts recovered in the
west that have been assigned, primarily by Herrmann, to Phrygian work-
shops. We have seen that Herrmann in his 1966 and 1975 contributions
attributed, incorrectly it was claimed, a number of bull attachments to Phry-
gian workshops. In his 1984 paper he has argued that it is possible to recog-
nize even more bronze Phrygian exports to the west, thereby considerably
increasing the inventory of exports, and at the same time expanding the
range of categories dedicated by Phrygians. Two of the objects purported
to be Phrygian derive from Samos, two from Olympia, and one each from
Delphi and Lindos.
The Samos objects consist of a winged siren-like attachment for a caul-
dron, and an ibex. The attachment was first published by Walter (1963: 294,
pl. 342a), and was assigned by him to Phrygia, an attribution accepted by
Herrmann (1984: 19f.). It is an odd piece indeed, not a typical siren attach-
ment of the oriental or Greek form, and it was recognized by Kopcke (1963:
284, no. 57) as a provincial work, probably of local manufacture.35
mention here the Olympia and Athens examples: probably as a reaction to Kyrieleis 1977, a
critique of Hermanns attributions.
35 Birmingham 1961: 190, mentioned siren figures excavated at Samos. Probably she was
discussing this (single) example. See also A.C. Brookes, Greek Siren Cauldron Attachments,
AthMitt 99, 1982: 608, who also mistakingly claims that siren attachments derive from Altin-
tepe. Birmingham (191) also calls the classic siren attachments Urartian-Phrygian which is a
meaningless term. In a recent article F. Isik (AnatStud 37, 1987: 163178) attempts to document
8th century bc cultural influences of Urutu on Phrygia, and to some extent, vice-versa: here,
for example (169), by claimingwithout documentationPhrygian influence on Urartian
painted pottery, which in fact is rare and usually dated post-8th century. Isik (169) considers
the siren cauldrons from Tumulus MM at Gordion to be Urartian exports (gegenstndlich
belegt), Unfortunately, Isik does not cite the extensive literature attributing these cauldrons
The second Samos object was also first published by Walter (1963: 294 f.,
pl. 342c), and also assigned by him to Phrygia; it is not mentioned by Herr-
mann. Jantzen (1972: 70, pl. 72, B1282) considered it to be Assyrian, but
Kopcke (1968: 291f., no. 119) correctly recognized its apparent Iranian (Berg-
lands) background (Muscarella 1973: 237).
In his 1984 paper Herrmann also considered an odd, crudely executed
attachment of a male figure with outstretched arms (1966: 32, 49, 81 f., pl. 20,
A15) and a Greek-like (to my eyes) siren attachment discovered in 1965 at
Olympia (1984: 19f., pl. 5, A15a) to be Phrygian productions. In 1966 (82 f.,
nos. 43, 54, pl. 27) he discussed a siren attachment from Delphi and one
from Lindos. For both he suggested that they might be East Greek, but he
also suggested that they could be Phrygian; in 1984 (18, 20) he rejected the
East Greek attribution and claimed them for Phrygia.
For all the above attachments Herrmann correctly noted that they erhe-
bliche stilistische Unterschiede aufweisen, but he urged that this fact is not
a problem, inasmuch as Phrygian art known in its homeland, aside from pot-
tery, is not homogeneous. Indeed, Herrmann is correct in this observation
about Phrygian art. But in a meaningful archaeological discourse it cannot
occur that any odd, unidentifiable, piece found in the west, artifacts that
have absolutely no parallels either in Phrygian art or among its excavated
artifact forms which description obtains for all the attachments discussed
here may casually be assigned to Phrygia. Such attributions do not lead to
historical reality.36
Summarizing in brief the historical and archaeological evidence for
Phrygian-Greek contact limited to the period of Midas-Mita, we have at
Gordion Greek pottery and fibulae, and probably a Greek princess; and in
Phrygia at large we have a Greek alphabet. In the Greek sphere there is
Midas throne at Delphi, and bronze fibulae, belts, probably some bowls, and
a cauldron,37 all dedications to various Greek deities at several sanctuaries.
From an analysis of the archaeological finds and of the historical sources we
have been able to recognize conscious knowledge of and friendly interac-
tion with Greek society, both on the west coast and the mainland. Among
to north Syria: he does cite Herrmann 1966 for the list of the Gordion cauldrons but he offers
no evidence that he has read Herrmann-who argued for a north Syrian origin.
36 See my discussion on the problem of Median art and the assignment of stray artifacts
to that category in JNES 46, 2, 1987: 109 ff., 126 f.

37 And clothing, if Boehmers conclusions are correctwhich we may never know given
the fragile nature of cloth.

the Greek sanctuaries the Heraion on Samos has in quantity and variety the
largest number of Phrygian dedications preserved: three or four fibulae, at
least one belt, and perhaps more, two bowls, and a cauldron; Paros has one
fibula; Lindos four or five fibulae; Olympia has fibulae and a bowl; Argos has
several fibulae and three bowls; Perachora has bowls. Surprisinglybecause
we want to have material here to inform us about other kinds of objects
besides a throne that may have been dedicatedwe have, to my knowledge,
no Phrygian material preserved at Delphi.
I do not think we can speculate at the present why Samos has more
material than that from the other sanctuaries. The reasons could be simple
ones, that Samos is closer to Phrygia and was easier for a Phrygian to visit;
or perhaps the extant material at other sanctuaries is merely a result of
accident, much having disappeared over the centuries. It is also possible that
here Hera was identified with Kybele, but we do not know.38
In the late 8th century bc Phrygia under the rule of its dynamic King
Midas was actively involved in international affairs. Midas intrigued and
sought alliances to the east and south with Urartu and Tabal and with
the north Syrian states, also with Assyria, first as an enemy and then after
709bc apparently as an ally. Equally, Midas involved Phrygia in political and
cultural affairs with the west, with the Greek states on the coast and with
the Greek mainland itself.
Phrygians travelled long distances overland and by sea to honor Greek
deities, and perhaps Greeks visited Gordion and other Phrygian centers.
King Midas, an international personality, and the Greeks were well ac-
quainted with each other.39 And the historical reality of this acquaintance
38 The Heraeum yielded dedications from Egypt, North Syria, Assyria, Iran, possibly the
Caucasus, and one object, a bell, from Urartu. The sanctuary was manifestly well known
to Near Easterners who must have had special reasons for dedicating votive material there.
Whether each people had the same reason is of course unknown to us.
39 Here we may mention the conclusions of K. DeVries in his provocative paper Greeks
and Phrygians in the Early Iron Age, in From Athens to Gordian, ed. K. DeVries, Philadelphia,
1980: 3350. DeVries argues that there was a correspondance, not fortuitous, between the
ways of life and certain attitudes of the Phrygian court society and that revealed in the
Homeric epics and he implies that the epics may actually reflect a living counterpart
in Phrygia (especially pages 40 ff.). This conclusion is exciting and stimulating: but much
of it is based on subjective archaeological and personal interpretationfor example, that
hundreds of not free women worked in the Terrace Building at Gordion, and that gift
giving may account for the presence in Tumulus MM of vessels bearing personal names.
Indeed, the Homeric epics could in part have been based on the megaron society at Gordion,
but there is no hard evidence that this was so, or that other such societies. Still unknown
must have informed the background to the Greeks continued fascination

with Midas over the centuries.
Bibliography
E. Akurgal
1955: Phrygische Kunst. Ankara.
1961: Die Kunst Anatoliens, Berlin.
1966: The Art of Greece, New York.
1983: Zur Datierung der ltesten ionischen Mnzen mit Lwenkopfdarstellun-
gen, in Beitrge zur Altertumskunde Kleinasiens, R.M. Boehmer, H. Haupt-
man editors, Mainz: 111.
M. Balmuth
1971: Remarks on the Appearance of the Earliest Coins, in Studies Presented to
George M.A. Hanfmann. D.G. Mitten, et al., eds., Cambridge.
R.D. Barnett
1967: Phrygia and the Peoples of Anatolia in the Iron Age, CAH fascicle.
J. Birmingham
1961: The Overland Route across Anatolia in the Eighth and Seventh Centuries
B.C., AnatStud 11: 186195.
K. Bittel
1942: Kleinasiatische Studien, Istanbul.
1950: Grundzge der Vor- und Frhgeschichte, Tbingen.
1970: Hattusha, New York.
Ch. Bunkenberg
1926: Fibules grecques et orientales, Copenhagen.
1931: Lindos: les petits objets, Berlin.
J. Boardman
1961: Ionian Bronze Belts, Anatolia 6: 179189.
1973: The Greeks Overseas, London.
R.M. Boehmer
1972: Die Kleinfunde von Bogazky, Berlin.
1973: Phrygische Prunkgewnder des 8. Jahrhunderts v. Chr. Herkunft und Ex-
port, AA: 149172.
to us may not have contributed their essence. At the same time, DeVries suggestions are
not to be eliminated from our thoughts, especially given the evidence presented in this
paper.
1983: Eine tabalische XII, 9 Fibel aus Assur, in Beitrge zur Altertumskunde
Kleinasiens, R.M. Boehmer, H. Hauptmann eds., Berlin: 7582.
S. Bulu
1979: Ankara Frig Nekropolnden Tmls Buluntulari, Habilitation Thesis,
Ankara.
E. Caner
1983: Fibeln in Anatolien PBFXIV/8, Munich.
A. de Ridder
1896: Catalogue des bronzes sur lAcropole dAthenes. Paris.
K. DeVries
1981: contributions in Young 1981.
I.M. Diakonoff, V.P. Neroznak

1985: Phrygian, New York.
T.J. Dunbabin
1957: The Greeks and their Eastern Neighbors, London.
J. Friedrich
1941: Phrygia, in PW, Stuttgart: 782891.
A. Frtwangler
1890: Die Bronzen und die brigen kleineren Funde von Olympia, IV, Berlin.
A. Goetze
1957: Kleinasien, Munich.
H.V. Herrmann
1966: Die Kessel derorientalisierenden Zeit, OL VI, Berlin.
1975: Hellas, in RLA Band IV: 303311.
1980: Kessel, in RLA Band V: 573:578.
1984: Kesselschmuck, nachtrge, AthMitt: 1733.
G. Huxley
1959: Titles of Midas, GrRomByzSt 2,2: 8599.
U. Jantzen
1972: gyptische und orientalische Bronzen aus dem Heraion von Samos, Samos
VIII, Bonn.
L.H. Jeffery
1976: Archaic Greece, London.
D. Kagan
1982: The Dates of the Earliest Coins, AJA 86, 3: 343360.
A. Knudsen
1964: From a Sardis Tomb , Berytus XV: 5969.
A. Koerte
1903: Agamemnon, in PW, Supp. 1, Stuttgart: 19.
1904: Gordion Ergebnisse der Ausgrabung im Jahre 1900, Berlin.
G. Kopcke
1968: Heraion von Samos: Die Kampagnen 1961/1965 in Sdtemenos, AthMitt 83:
250314.
W. Kroll
1932: Midas, in PW, Stuttgart: 15261540.
H. Kyrieleis
1977: Stierprotomen orientalisch oder griechisch?, AthMitt 92: 7189.
M. Le Jeune
1969: Discussions sur lalphabet phrygien, SMEA 10: 1947.
D.D. Luckenbill
1927: Ancient Records of Assyria and Babylon I, Chicago.
M. Mellink
1965: Mita, Mushki and Phrygians, Anadolu Arastirmalan 2: 317325.
1981: contributions in Young 1981.
O.W. Muscarella
1967: Phrygian Fibulae from Gordion, London.
1968: Winged Bull Cauldron Attachments from Iran, MMA Jour 1: 718.
1973: review of Jantzen 1972 in AJA 77, 2: 236237.
1977: The Archaeological Evidence for relations Between Greece and Iran in the
First Millennium B.C., JANES 9: 3357.
1982: review of Young 1981 in Quarterly Review of Archaelogy 3, 4: 710.
1988: The Background to the Phrygian Bronze Industry, in Bronzeworking Cen-
tres of Western Asia 1000539 B.C., J. Curtis ed., London: 177192.
P. Neve
1972: Bericht ber die Deutsche Bogazky-Expedition im Jahre 1970, TurkAD 19,
1: 175186.
H. Payne
1940: Perachora I, Oxford.
P. Perdrizet
1908: Fouilles de Delphes, V, Paris.
J.N. Postgate
1973: Assyrian Texts and Fragments, Iraq 35, 1: 1336.
F. Prayon
1987: Phrygische Plastik, Tbingen.
L. Roller
1983: The Legend of Midas, Classical Antiquity 2,2: 299313.
K. Sams
1974: Phrygian Painted Animals: Anatolian Orientalizing Art, AnatStud 24: 169
196.
1978: Schools of Geometric Painting in Early Iron Age Anatolia, in Proc. of the
Xth International Congress of Classsical Archaeology, Ankara: 227236.
1979: Patterns of Trade in First Millennium Gordion, Arch. News 8, 2/3: 4553.
E. Simpson
1983: Reconstructing an Ancient Table , Expedition 25, 4: 1126.
1986: Royal Wooden Furniture from Gordion, Archaeology 39.6: 4047.
A. Snodgrass
1971: The Dark Age of Greece, Edinburgh.
M. van Loon
1966: Urartian Art, Istanbul.
H.T. Wade-Gery
1952: The Poet of the Iliad, Cambridge.
G. Waldstein
1905: The Argive Heraeum, New York.
H. Walter
1963: Die Ausgrabung irn Heraion von Samos (19521962), ADelt 18, 2: 286296.
H. Winckler
1901: Altorietalische Forschungen, 2nd series, Leipzig.
R.S. Young
1958: The Gordion Campaign of 1957, AJA 62, 2: 139154.
1960: The Gordion Campaign of 1959, AJA 64, 3: 227243.
1963: Gordion on the Royal Road, ProcAmerPhilSoc 107, 4: 348364.
1969: Old Phrygian Inscriptions from Gordion, Hesperia 38, 2: 252296.
1981: Three Great Early Tumuli, Philadelphia.
chapter twenty-four
GREEK AND ORIENTAL CAULDRON ATTACHMENTS: A REVIEW *
Plates IVb
To Hans-Volkmar Herrmann with Respect
The 8th century bc was truly one of the major periods in the history of
the western world. During this century, primarily in its second half, the
Greeks, fully emerged from a Dark Age, commenced active trade and multi-
level cultural exchange with the Near East, the Orient. The evidence for
these activities is primarily the creation of the Greek alphabet, and also the
presence of manifold oriental objects and motifs recovered in Greek soil.
One of the most prevalent classes of oriental material that reached Greece
in the 8th century bc was the large bronze cauldron fitted with winged
siren1 and winged bull-head attachments. More significant in a cultural
sense is the fact that the imported cauldrons and attachments were copied
by local craftsmen. While not alone among the imported material copied
and adapted to satisfy local needs, the Greek-made cauldrons, especially
those with sirens, constitute some of the best evidence available of the phe-
nomenon known as orientalizing.
Cauldrons with attachments have been of special interest to me since
1957, when I saw my first siren and bull cauldrons in the great Tumu-
lus MM at Gordion (pl. Ia), which encounter led to my 1962 paper on sirens.
Although much has been written about cauldrons since 1962 (especially
Herrmanns potent and laudable monograph, 1966a), some of the ques-
tions asked remain unanswered, some questions were never asked, and
conclusions about their origins are still contested. It seems appropriate,
* This chapter originally appeared as Greek and Oriental Cauldron Attachments: A
Review, in Greece between East and West: 10th8th Centuries B.C., eds. G. Kopcke and I. Toku-
maru (Mainz: von Zabern, 1992), 1645.
1 I will call them sirens as a convenient, well-known term: see Amandry 1956, 244, n. 14;
Herrmann 1966a, 28; Wartke 1985, 90. In this study I shall not cite or discuss individual
differences among the sirens or their identification and background, as these matters are
discussed elsewhere. I wish to thank Susan Pattullo for reading and commenting upon this
paper.
726 chapter twenty-four
therefore, to present an overview of recent scholarship and thoughts on

these remarkable objects.
The present paper will comment upon some of the crucial issues that
have occupied cauldron researchers both before and after the early 1960s,
although here I shall concentrate on the work of the last 30 years. These
issues include: the number of both oriental and Greek-made cauldrons
and stylistic determinations of the two forms; their actual findspots; the
oriental origins and scholarly (and non-scholarly) views on this question;
chronology; their function in Greece and the Orient; and the problem of
determining the process of their movement to Greek soil.
Siren Cauldron Attachments
In the extensive literature on siren attachments recovered in the Greek

sphere discussions have focused on questions that to a large degree reflect
Greek archaeological interests: what culture(s) in the Near East (Orient)
produced and exported them; which of the examples recovered is an orien-
tal import, which a local production; and what is the exact number of sirens
in existence? We should also add, how many cauldrons are involved and how
many were actuallynot putativelyexcavated? A review of these ques-
tions follows, beginning with the number of sirens, both Greek and oriental,
in the corpus.
In 1931 two catalogue listings of siren attachments appeared: Ipsen (apud
Lehmann-Haupt 1931, Anmerkungen 21f.), and Kunze 1931, 267 ff. The former,
little known and barely cited (but v. Herrmann 1966a, 28, n. 4), listed 43
examples, while the latter, which became the seminal catalogue, listed 50
examples. Not included in Ipsens list are Kunzes nos. 11, 15, 22, 31, 39, 42
and 45. In 1950 Kunze raised the number to 59 (1950, 100 f.). He removed
his 1931 no. 11, a terracotta head acquired in Crete by Arthur Evans (now in
the Ashmolean Museum), because he came to believe that it was not in fact
a siren head (infra and n. 10). He added six more examples from Olympia,
one from Delos, an example in the British Museum, and two objects from
Argos, one of which is a siren tail, the other a bull figure (infra), Herrmann
(1966a, 28f., n. 8, 187) removed four items from Kunzes 1931 list, nos. 11,
20 (a tail that joined 01 AI), and nos. 21 and 39, because they were not
considered to be sirens.2 He added more recent finds from Olympia, and the
2 On May 11, 1959, Kunze told me he wished to remove no. 21 from his list; no. 39 is
technically a siren but of a later type (Kunze 1931, 278 f.).

greek and oriental cauldron attachments: a review 727
eight from the royal tomb at Gordion, Tumulus MM (pl. Ia; infra), arriving
at a total figure of 78 examples in the corpus.3 This figure did not include
two aberrant attachments from Samos, which Herrmann believed did not
correctly belong to the class of sirens (29, 147, 187).4
Jantzen (1967, 91ff.) argued for the addition of the Samian pieces, recog-
nizing them as local, orientalizing productions of an oriental form. In his
review of Herrmann 1966a, Amandry (1969, 797) also accepted the Samian
pair as legitimate members of the class (see also Rolley 1984, 278); he further
added another example from Delphi (Rolley 1967, no. 140; pl. Ib) and the
four aberrant hammered sirens from Salamis on Cyprus (Karageorghis 1973,
97ff., figs. 1823, pl. 244), accepting them as local concepts of the class. At
the same time, Amandry removed from the corpus Kunzes and Herrmanns
nos. 7, 8, and 9 (of the alleged Van group), maintaining that these pieces have
never been seen and, in fact, may never have existed. These additions and
subtractions yielded a total number of 82.5
Herrmann in 1984 (21) added a pair on an unexcavated cauldron in
Munich (infra; pl. IIa), but he provided here no total figure. Barnett (1986,
112) noted Amandrys 1969 deletions, the additions of the Samian pair and
the one from Delphi, but not the four from Cyprus; he also cited the recently
published siren in the Glencairn collection in Bryn Athyn, Pennsylvania
(Romano and Pigott 1983)6 but not the Munich cauldron. Although his fig-
ures add up to 80, Barnett claimed 81 sirens for the corpus.7 Finally, Rolley
(1984, 280) added three more sirens (said to be Greek) from Delphi (unpub-
lished).8
My review of the figures yields the following results. Appropriately start-
ing with Herrmanns 1966a figure of 78, I would remove his no. 51, a bull figure
3 See the Nachtrge zu S. 56 ff. on 187 for a correct count.

4 Note that Herrmann mentioned but one siren from Samos. In 1989, 341, n. 35, I repeated
this error, neglecting Jantzen 1967, and I challenged Judy Birmingham on the number of sirens
on Samos: she, of course, was correct.
5 In 1978, 62, n. 2, I gave the low figure of 79 following received lists.
6 The authors note (126) that this piece is close to Istanbul no. 41, while Amandry 1969,
799, thinks that no. 41 is close to the Gordion examples: I accept the former but not the latter
conclusion.
7 Note that Barnett misstated that Herrmann catalogued 78 oriental sirens: this figure, of
course, included both Greek and oriental examples. Barnett also misstated the actual number
of sirens excavated in Greece (see infra).
8 In Filippakis et al. 1983, 131, sample nos. 76 and 77 are listed as oriental sirens from
Delphi, and on 132 sample no. 78 as Greek: if the first two are the same as those called Greek
in 1984 by Rolley, the Greek and oriental lists presented in this study would be skewed. Future
publication may resolve the apparent problem.
from Argos (pl. Vcd, Kunzes 1950 addition; see Herrmann 1966a, 79 f., 153,
156), because I find no clear evidence that a siren was necessarily originally
present. Kunze (1931, 12), Amandry (1956, 245, n. 17), and Rolley (1984, 283)
envisioned a siren with a bull placed on each wing, the latter ingeniously
citing as an exact formal parallel the cauldron from Salamis, which has
a griffin on each wing of a siren (eight griffins with four sirens). To me
the case is not proven.9 Following Amandrys suggestion, nos. 7, 8 and 9
should be removed (but to be re-added if ever they surface). This leaves
74 examples. To this figure we should add the two examples from Samos,
the four from Delphi cited by Rolley (1967, 1984), the four from Cyprus
(recognizing their uniqueness), the one in the Glencairn collection, and
parenthetically the clay head in the Ashmolean Museum.10 We are also
obliged to add, tentatively accepting them to be genuine, the two sirens on
the cauldron that surfaced on the Munich art market (pl. IIa).
These figures add up to 87 bronze sirens in the corpus,11 88 sirens, if we
accept the Ashmolean Museum clay head. A number are clearly aberrant,
non-canonical, but clearly (or seemingly) expressing the same form and
concept: those from Samos and Cyprus, two from Olympia, A14 and A15
(Herrmann 1966a, 32; idem 1984, 17ff.), and the two on the Munich art market
cauldron.
9 Blegen 1939, 430, stated that there are traces apparently of a hand on the plinths
base; Kunze 1950, 96 says that on the object ruht eine Hand flach auf. I see in the photo
a raised area to the bulls right and a projection: a thumb? Rolley 1984, 283, cites a similar
example from Delphi but did not supply an illustration. In any event, no siren with bulls
on its wings exists. Gabelmann 1964, 6, 8, accepts Kunzes conclusion and goes further: he
believes that a bronze lion figure and a siren in the Petit Palais (Herrmann 1966a, 155, no. 32,
59, no. 62) belong to the same cauldron and thus represent another example of animal-siren
juxtaposition: which to him is Urartian. Each of these assertions is unacceptable: the Petit
Palais collection is a modern, not an ancient, assemblage. Ingrid Strm in her paper also
accepts the presence of a siren (infra, 52). And since the sirens come from North Syria (infra),
she further concludes that the juxtaposed animal handle, and other handles (Kesseltiere),
are also from North Syria (54, n. 29): but there is no archaeological or stylistic evidence to
support this conclusion (Muscarella 1970, 115).
10 Kunze 1950, 101, removed the head from his 1931 list, but later (apud Herrmann 1966a,
28, n. 8) reversed himself again. In CVA Oxford fasc. 2 (1931), 55, pl. 2, 24, it is called a siren
head. Levi 1945, 290, said it belongs to a vase; Boardman 1961, 60, no. 258, also calls it a siren
head; on 52 he cited Kunze no. 11 and the 1950 retraction. Amandry and Barnett implicitly
accept Herrmanns rejection of the head; in any event, it is not a bronze siren attachment.
11 After the completion of this paper I was informed that a siren tail had been excavated at
Isthmia; it will be published by I. and A. Raubitschek. This brings the corpus up to 88 bronze
sirens, 89 for the full corpus.
How many of these sirens were excavated, and where? Concerning ourselves
first with the oriental sirens recovered in the Greek sphere, we recall that
Herrmann listed 21 examples from Olympia, two of which were recovered on
a cauldron with protomes (1966a, 11ff., 30f., 57, 187, pl. 1; idem 1984 mentions
20). Agreeing with Brandes (apud Herrmann 1966a, 55, n. 32, 83, n. 24),
Brookes (1982, 609f.), and Rolley (1984, 278), I would remove Herrmanns
nos. A14, A15, and A15a (Herrmann 1966a, 187; Herrmann 1984; Taf. 5: 13)
from the oriental list and transfer them to the Greek category. The hair form
and herring-bone pattern of A14 are not the same as those of any oriental
siren, and the fact that it is aberrant does not make it oriental (cf. its round
base to Herrmanns pl. 46, 1). A15 and A15atwo distinct pieces despite the
similar numbers assignedfit into the same category: the arms and hair
on both are not obvious oriental forms.12 This leaves 18 oriental sirens from
Olympia, and an estimated maximum number of 15 cauldrons (accepting
A3, A4, A5, A5a, and A10, A11 as pairs).
From Herrmanns Delphi list (57f.) of 13 examples, two of which were
recovered on a cauldron I would omit no. 37, the well-known bearded siren
in Copenhagen (pl. IIb); purchased from a dealer, it has no provenance,
Delphi or elsewhere (infra).13 Agreeing with Brookes (op. cit.) that it is not
oriental, I omit no. 44, in the Louvre. For the same reason I would omit
nos. 42, 43, 45, and 47 (as Brandes, op. cit.). Only at first view may no. 43
deceive, but the hair and face form and, especially, the attenuated and non-
articulated hands and arms against a short wing do not point very far east
(see also Rolley 1984, 278). No. 45 was described by Perdrizet (1908, 81) as
having no arms (but is considered to be oriental by Filippakis et al. 1983,
131, sample no. 21). No. 47 is a thin fragement with no head preserved.
After adding Rolleys 1967 (no. 140) example, we arrive at the figure 8 for
the number of oriental sirens from Delphi, with an estimated maximum
number of seven cauldrons.14
There are two oriental sirens on a cauldron from the Ptoion. The two
single, detached examples from Argos and Delos (nos. 52, 53, 50 and 55)
equal two cauldrons. A single example from Lindos on Rhodes (no. 54), one
cauldron, is not easy to attribute (especially lacking autopsy). Originally,
12 Kyrieleis 1967, 7, claimed A14 was Urartian; Herrmann 1984, 20, claimed A15 and A15a
to be Phrygian; to Kilian-Dirlmeier 1985, 247, A15 is Syrian, A15a Phrygian; see also Herrmann
1966a, 83, n. 24.
13 It was once offered to the Metropolitan Museum of Art for sale by a Dr. Hirsch.
14 Note that I accept no. 39 as oriental, contra Brandes; see also n. 8 above.
Herrmann (1966a, 82f.), who recognized the difficulties of classification,

thought it (and no. 43) to be East Greek15 or Phrygian, but in 1984 (20) to
be Phrygian; Jantzen (1967, 93, n. 11) and Rolley (1984, 278) considered it to
be East Greek. To my eyes it is either oriental or a closely copied Greek/East
Greek work. Hence, I leave its attribution as uncertain.16 The six examples,
two each on three cauldrons together with protomes, from Italy, Vetulonia
and Praeneste (nos. 5661) are of course oriental.
This reexamination indicates that at least 30 oriental sirensas opposed
to Herrmanns 38have been excavated in the Greek sphere31 if the
Lindos example is accepted as oriental; these figures represent an estimated
maximum number of 25 imported oriental cauldrons (26 with Lindos). But
note: if it is accepted that the three Italian cauldrons reached Italy via Greece
(Herrmann 1966a, 86), it would accordingly follow that there is evidence
for 36/37 oriental sirens on an estimated 28/29 cauldrons that originally
reached Greece (see Griffin Protomes, infra).
Kunze (1931, 274), Herrmann (1966a, 4, 87, n. 13, 106 f., 143, n. 4, 145 ff.; 1979,
137, 142), and Jantzen (1967, 91) maintained that some oriental cauldrons
were imported into Greece with two siren attachments (since only pairs
occur in the West) and dedicated without the addition of griffin protomes,
e.g., Ptoon and Delphi. Others, rare, had the addition, e.g., Olympia B 4224,
an example from Delphi (Perdrizet 1908, 85, fig. 289), and those from Italy.
Herrmann is unsure whether the Greek cauldrons also included protomes,
while Rolley (1984, 280f.) noted that Greek sirens in Greece were recovered
separated from their cauldrons. Rolley also believes that the Ptoon and prob-
ably the Delphi oriental cauldrons once had protomes. I have no contribu-
tion to make on this issue (v. Muscarella 1981a, 50 f.), except to note that both
arguments assume there were two sirens per imported cauldron, a position
I have tentatively accepted here with regard for cauldron estimates (see the
discussion on the Munich art market cauldron, infra). However, given the
evidence from Gordion (pl. Ia), where there are four sirens on each of two
cauldrons, we are constrained from assuming that no oriental cauldron with
four sirens reached Greece (see Herrmann 1966a, 145).
Turning to the excavated Greek-made sirens, Herrmann (1966a, 82 f., 102;

1984, 20, n. 19; also Jantzen 1967, 93, and Weber 1974, 34) listed 12 from the
15 Herrmann 1966a, 82, believed that nos. 43 and 54 were made in the same workshop,
which I do not accept; see also 46, n. 36 and 55, n. 32.

16 Agreeing with Herrmanns (84) uncertainty; he sees no. 43 and the Lindos siren to be
Anhang bei der orientalischen Gruppe.

mainland. Of the five listed from Olympia (91f., 102), only three were actually
excavated there, nos. 2 and 3, a pair, and 10. But no. 11, in Boston, is clearly a
mate to no. 10 (Herrmann 1966a, pls. 32, 33), and is thereby rightly included
with the Olympia finds. No. 6 in the British Museum, however, is a purchased
piece with no provenance (pace Herrmann 1966a, 92, 95, n. 8).17 To the four
excavated examples I add the three removed from the oriental list (nos.
A14, A15, A15a), for a total of seven Greek-made sirens from Olympia, which
equals five cauldrons.
To Herrmanns list of two Greek sirens from Delphi, nos. 1 and 4, I add
the five removed from the oriental list (nos. 42, 43, 44, 45 and 47)18 and the
three reported by Rolley 1984, for a total of 10 Greek sirens from Delphi,19 and
about 810 cauldrons. Athens has four sirens on 4 cauldrons, nos. 5, 7, 8 and
9?, not five as Herrmann claims (1966a, 29, n. 9, 102, 113), for his no. 12, in
Munich, is not an excavated piece (v. Kunze 1931, no. 43). Herrmanns no. 9
is a problem piece, for only a head exists and in a poor state of preservation.
Nevertheless, it may be a siren as claimed and is included here tentatively
pending autopsy. And Samos has two examples on one cauldron.
We arrive thus at the figure 23 as a base number of excavated Greek-made
sirens from the Greek sphere: 21 from the mainland and 2 from Samos. If one
considers the Lindos siren to be Greek/East Greek, the figure is 24, and the
oriental figure would be reduced to 30. But it would be too harsh a position
to defend if we were to exclude from the above list Herrmanns 1966a, nos. 6
and 12, from two cauldrons, in London and Munich. Both are Greek-made
and surely derivedif not scientifically excavatedfrom somewhere in
the Greek mainland. And although the clay Ashmolean head from Crete is
not unambiguously a clear-cut member of the siren class, I include it here
tentatively, as possibly a fragment of a siren.20
The total count of Greek-made sirens is thus 26, from about 2123 caul-
drons, a figure more than double the number hitherto recognized. These
figures do not include the four locally made sirens from Salamis on Cyprus,
for this region warrants a distinct cultural designation.
17 It was purchased in 1914 from a Greek who lived in Paris; it was he who said it came
from Olympia (information from John Curtis).

18 No. 44 was given to the Louvre as from Delphi by Perdrizet; I therefore accept its
attribution. Neither Amandry nor Herrmann cite Kunzes no. 31, hence I do not include it
(what and where is it?).
19 See Filippakis et al. 1983, 132, sample no. 78 for, apparently, one of these sirens.
20 On Crete griffin head protomes were locally made in clay, Hanfmann 1957, 243; Board-
man 1980, 60 f.; Blome 1988, 559 ff.

There are still only nine excavated sirens, from two sites, in the Near East:
four each on two cauldrons from Tumulus MM at Gordion (pl. Ia), and an
isolated example from Alishar (Verachram), a site on the Iranian side of the
Araxes River (Herrmanns nos. 1018).21 No examples have been excavated in
the heartland of Urartu (infra).
To summarize: 76 of the 8822 sirens in the corpus were excavated (in the
Near East, Cyprus, Greece, Italy), about 86 %. Of the corpus, 57 (30 oriental,
26 Greek, and Lindos), ca. 64%, derived from the Greek sphere (the Italian
examples are not included here; see supra), which is the same percentage
recorded by Ipsen and Herrmann; and of these 57 examples, representing
an estimated 47 to 49 cauldrons, about 45% are of Greek manufacture.
Olympia and Delphi have both oriental and locally-made cauldrons; Ptoion,
Argos, and Delos preserve only imports; and Athens and Samos preserve
only locally made examples. Significant is the 45 % figure: it indicates both a
more intense degree of local copying, that is, of orientalizing, and, perhaps
a larger number of Greek workshops in existence in the late 8th and early
7th centuries bc than previous studies had suggested.
Over a period of decades a number of scholars have attempted to identify

the specific oriental region(s) that produced the siren cauldrons and subse-
quently released them for export to the West (bibliography: Muscarella 1962,
321f., n. 14; idem 1970, 110, nn. 15, 16; idem 1978, 62, nn. 3, 5; Herrmann 1966a,
27f., 50f., 55ff.; idem 1966b, 118ff.; Wartke 1985, 93). There are essentially
three main opinions: one postulates a North Syrian, another an Urartian
main center, both accepting one or two examples to be of Assyrian man-
ufacture; the third, a minority opinion, argues for both a North Syrian and
an Urartian center, also accepting one or two examples to be Assyrian.
I still hold the view presented in previous writings (1962; 1970; 1978;
1988b), in agreement with Herrmann (1966a, b) and Wartke (1985), that
there is no evidence to indicate that even one siren cauldron was manu-
factured in Urartu. Greater North Syria23 still remains the best candidate for
21 Amandry 1969, 798, joins Herrmanns nos. 2 and 3 in Istanbul with no. 62 in Paris,
suggesting that they with a missing siren were originally on one cauldron as at Gordion.
Herrmann (1966a, 6 f., 77, 87, n. 13, 143, n. 2) previously had joined no. 62 to the Barberini
cauldron.
22 I include in these summaries the Ashmolean clay head and the London and Munich
sirens (supra). See n. 11 for another (Greek? oriental?) bronze example.

23 For a restricted boundary of the ancient North Syrian states see Winter 1988, 193f. She
also accepts a North Syrian origin (198, 204).

the region where the primary workshops flourished24 (see also Strms paper,
infra, 54). That a number of workshops functioned there is evidenced by the
variety in details of the sirens, although we are incapable of matching indi-
vidual sirens to specific North Syrian centers (Herrmann 1966a, 73).25
Unfortunately, some scholars have been misled to believe that a number
of sirens derived from Urartu and, while accepting a primary North Syrian
involvement, have argued that Urartu must have had at least a partial role in
their manufacture (e.g., Brandes apud Herrmann 1966a, 55; van Loon 1966,
110; Wfler 1975, 225, n. 1298; Amandry 1969, 799; idem 1978, 5 f.; Boardman
1980, 65). But in fact, only the example from Alishar in northwestern Iran
derives from Urartu-controlled land, and its presence there represents a
rare import from the west.26 The Berlin Toprakkale siren, long assumed to
have been excavated, is now known to have been purchased in Europe (not
in Van, as Piotrovskii 1967, 37). Consequently, fehlt damit die wesentliche
Sttze fr die urartische Provenienz der gesamten Gruppe (Wartke 1985,
87, nn. 2 and 3, 92; see also Muscarella 1988b, 28 f., n. 4; also Herrmann 1966a,
59, n. 36).
Notwithstanding the archaeological and art historical evidence pre-
sented since the early 1960s against an Urartian center, it is still asserted that
siren cauldrons were made in Urartu and exported west,27 through Anatolia
and into the Greek sphere (e.g., Rolley 1967, 15; idem 1984, 286; Karageorghis
24 One must think also of Riemschneider 1959, 66, who in an odd article was perhaps the
first scholar to argue against a major Urartian center; see also Riemschneider 1965, 105f.; Ipsen
(apud Lehmann-Haupt 1931, 495 f.) also cited North Syrian elements; so did Lehmann-Haupt
(497), but he claimed an Urartian origin for the sirens. Note that Kilian-Dirlmeiers article
(1985) on foreign material excavated in Greek sanctuaries should be used with caution. Her
list of the individual objects recovered (244 ff., and in charts and maps) is of value, but her
assignment of them to specific oriental cultures is not always accurate; the number of objects
claimed to derive from each culture as well as their percentages within the totality of finds
in the sanctuaries will have to be revised.
25 Herrmann 1966a, 66 f., does of course cite parallels between the sirens and specific
North Syrian reliefs, but he holds back from claiming that siren workshops existed in the
centers with the parallels. See also Winter 1988, 198, but she gives more emphasis to this
matter than Herrmann himself did.
26 Note that Riemschneider 1965, 105 f. and Amandry 1969, 798 and 1978, 5 (both of whom
believed that some sirens came from Urartu) thought it possible that they arrived there as
imports.
27 Part of the problem facing classical and other scholars is that they had been led
to believe, incorrectly, that Urartu conquered North Syria. But Urartu only reached the
northeastern area of Malatya: none of the major centers was ever conquered by Urartu. Note,
e.g., Barnett 1969, 146, who claims that Tell Rifa"at is in Urartu! See Dyson and Muscarella
1989, 19 and n. 93.
1973, 113, n. 9; Barnett 1982, 367; idem 1986, 112 ff., although here implying
some role for North Syria; Thimme 1982, 134; Zimansky 1985, 97; Isik 1987,
169).28
In the 1970s a cauldron with two siren attachments and a tripod stand
surfaced on the ever expanding Munich art market (pl. IIa; Muscarella 1978,
62, n. 2; Kellner 1976, 74, pl. 4, no. 99; Kroll 1979, 1980, 46 f., fig. 38). The
original publications of the cauldron were presented in exhibition cata-
logues of alleged Urartian art and artifacts consisting of recently acquired
objects in the possession of museums, private collections, and apparently
dealers.29 Along with other objects of unknown background the cauldron
was baptized Urartian by those who organized the exhibition.30 The fact
that the cauldron has no archaeological provenance (pace the label Prov.
Transkaukasien-Urartisch), and that the sirens cannot by any manipula-
tion of manifestly Urartian art styles be attributed to an Urartian workshop
were totally ignoredan old story:31 but one we shall ignore.32
The cauldrons siren heads differ in execution from all the others known:
the tilt of the head, the hunched neck and shoulders, the unlidded eyes and
short brows, the cursory execution of the hairform and lines, the slit mouth,
and the circular pattern on the chest and shoulder. The wing and tail and the
position of the arms and hands look oriental,33 the face less so.34 Although
puzzling, the Munich sirens appear (without autopsy) to be genuinen; one
would welcome a disinterested laboratory analysis.
But aside from the question of the cauldrons origin and the uniqueness
of the sirens, there is another problem to be addressed. Only two sirens are
28 See also the confused article, Korti-Konti 1971, 281ff. Also, Frankenstein 1979, 269, 271.
29 Kellners no. 92, a vessel, is listed as belonging to one G.N. In the Ghent catalogue it
turns up as no. 194, there listed as belonging to a Swiss private collector. Does this mean it was
sold after its exhibition in Munich? Kellner 1976, 10, laments the recent plunder of Urartu
eine zweite Zerstrung Urartus, seiner Kultur und Geschichte; and he calls his museums
exhibition ein Abschied von einer vorgeschichtlichen Kultur : but he ingeniously refrains
from discussing the Munich museums role in the Zerstrung and Abschied.
30 St. Kroll has since rejected the attribution of the Munich cauldron to Urartu (personal
communication).
31 See Herrmann 1966b, 124 f., n. 148. It seems over the past years that if an object arrives
in Munich from a plundered site it is considered to be an Urartian artifact: which makes

Munich the River Jordan of Urartian art. Indeed, many Urartian objects do reach Munich,
in a continuous stream.
32 See infra, the discussion on alleged Urartian bull cauldrons published in the same
exhibition catalogues on Urartian art.

33 The hands have six fingers, a good sign: v. Barnett 1986.
34 One could play the East Greek card, but this would create only more confusion. It clearly
was made in a workshop other than those that made the other sirens.
present, which feature raises the question: can an unexcavated, unprove-

nanced cauldron be invoked as proof that cauldrons with two sirens, and
no protomes, existed in the Orient, cauldrons which were exported in this
manner to the Greek sphere?35 The question remains pertinent even if
we realize that we have no knowledge whence the Munich cauldron was
plunderedin Anatolia, North Syria, Greece, or elsewhere. It exists with-
out a history.
The Munich cauldron has also been cited to demonstrate an important
conclusion. Thimme (1982, 132) believes he has decisively eliminated Herr-
manns wiederholt vertretene These that the sirens are not Urartian. He
notes the similarities that link the tripod under the Munich cauldron, the
tripod under an unexcavated, but certainly Urartian, bull cauldron in Karl-
sruhe (infra), and the tripod of the Urartian bull cauldron from Antilepe. To
him, these parallels prove the Urartian origin of the sirens; the North Syrian
These is endgltig widerlegt. Missing from this proof is reference to the
style of the sirens and their archaeological history, as well as awareness that
bull-feet terminals are present on tripods in many areas of the ancient Near
East, where they occur mainly on non-Urartian objects.36 That the Munich
siren cauldron (assuming the tripod is an original part of the ensemble),
and the Altintepe and Karlsruhe bull cauldrons have the same tripod form
is striking, but the shared feature alone cannot carry the weight of the proof
Thimme imposes. Other possibilities, such as shared technological knowl-
edge, may have played a role.37
Since the publication of Muscarella 1962 and Herrmann 1966a and b, the
only work that systematically argued against a North Syrian origin for the
sirens and presented a vigorous defense of the Urartu-These is Kyrieleis
1967.38 Notwithstanding the earnestness of his article, I find Kyrieleis
Urartian parallels for the sirens unconvincing. His perplexing citation of
non-Urartian objects, some unexcavated, which he asserts are Urartian, is
completely unconvincing (see Muscarella 1970, 126, n. 20, and Wartke 1985,
35 Which is a position I tentatively hold: see Muscarella 1981a, 50f. and supra.
36 I give but a few examples: Muscarella 1980, 56 f., nos. 119, 172, 177; Muscarella 1970, 120,
fig. 11; van den Berghe and De Meyer 19821983, nos. 183, 184 (see n. 39 below); Kyrieleis 1969,
76ff.; Kohlmeyer and Saherwala 1984, fig. 19; Herrmann 1966a, 4, fig. 6; Layard 1853, figs. on
178 f.; see also Liepmann 1968, 55 f.
37 This argument cannot be applied to the production of sirens, at the very least because of
style. One would appreciate disinterested laboratory analysis to demonstrate that the tripod
belongs to the cauldron.
38 Barnett 1986, 113, mistakenly placed Kyrieleis among the North Syrian advocates; he also
seems to imply that Herrmann believed that Urartian cauldrons with sirens reached Greece.
95, 96, n. 53, 98).39 A few more views are worth mentioning for the record.
Wfler (1975, 254, n. 1298) completely excludes North Syria as the Heimat
and posits that they were made in Tabal and/or Phrygia, Urartu (the Berlin
Toprakkale example), and Assyria; and Calmeyer (1973, 130) excludes both
North Syria and Urartu, positing an origin in east Anatolia.
The sirens believed to be of Assyrian manufacture are an example in the
British Museum and the one in Copenhagen (Herrmanns nos. 1, 37, pls. 22,
23; supra).40 Beginning with Kunze (1931, 271, no. 1), the London siren, with
its round, plump face and chest, has rightly been singled out as standing
outside the canonical forms of the corpus and is assumed to have derived
from Assyria (van Loon 1966, 110; Kyrieleis 1967, 12 f., n. 68; Amandry 1969,
798; Muscarella 1970, 111; Rolley 1984, 278).41 Herrmann (1966a, 63 f., 148;
1966b, 119; 1980, 575) believed it to be the prototype from which the North
Syrian group developed, an assertion rightly rejected by Kyrieleis (ibid, 12 f.)
and Amandry (ibid).42
The Copenhagen head (pl. IIb) also stands out from the canon in the style
of the head and beard, the wing and tail patterning, and in the position of
the hands, which hold onto the rim of the cauldron. Kunze also thought
this attachment to be Assyrian, and most scholarsHerrmann, Kyrieleis,
Rolley, even Akurgal (1968, 38)have followed him; only Riemschneider
(1959, 66) and Amandry (but not decisively1969, 798) believed this siren
to be Urartian. I do not see in this head an Assyrian background but rather
a product of a North Syrian workshop (1962, 327, pl. 104b; 1970, 126, n. 17).
The lack of a mustache (at least it is not apparent or prominent), the hair
configuration and the spiral volute on the tail point to North Syria, from a
workshop that apparently was aware of Assyrian art.43
39 See also Seidl 1988, 172 f., n. 1. But Kyrieleis 1967, 13, n. 68 is correct in rejecting an
Urartian attribution for the Erlangen tripod statuette; see also Moorey 1973, 83ff., 90. Moorey
(84 f.) also cogently relates the face and hair of the Erlangen statuette to that of the British
Museum siren, infra; cf. Herrmann 1966a, 66.
40 Both exhibit ears, which are not found on other Near Eastern examples; some Greek
examples have ears, others do not.

41 For excavated Assyrian objects in the Greek sphere see Muscarella 1977, 48, n. 67; add
Brker-Klahn 1973, 51 ff. pl. 26.

42 Lehmann-Haupt 1931, 494 believed it was excavated in Nineveh, arriving there in antiq-
uity from North Syria, but that it and other sirens were of Urartian manufacture. It follows
that the Assyrians copied at least one North Syrian cauldron.
43 I call attention here also to an unexcavated bucket with six bearded, divine siren-figures
(Muscarella 1981b, 269 f.), which on stylistic grounds I attributed to a North Syrian workshop,
and which I compared to the canonical siren attachments; see n. 70.
Winged Bull Cauldron Attachments
The class of winged bull head attachments excavated in the Near East can be
divided into two main groups based on stylistic details and manufacturing
techniques. One can recognize a homogeneous Urartian group and a general
Near Eastern group that is considerably diversified, but in which we may
discern two or more cultural backgrounds (Muscarella 1968, 10 ff.; 1970, 111 ff.;
1989, 262f.). The Urartian and Near Eastern groups are quite distinct, yet
paradoxically many scholars have failed to observe the distinctions.
Thirteen examples of the Urartian group have been excavated: four each
on cauldrons from Altintepe, Toprakkale,44 and Guschi, the last two convinc-
ingly reconstructed from the incomplete records; and an isolated example
from Alishar, where, as reported above, a siren was also recovered (Amandry
1956; Hanfmann 1957; Herrmann 1966a, 118, n. 10; Piotrovskii 1967, 36 ff.; Mus-
carella 1968, 7ff., n. 2).45 Not a single example of an Urartian bull attachment
has been excavated outside of Urartian territory. This geographic unity par-
allels the groups distinct stylistic characteristics, which include a separate
casting of the head and the wing-tail unit, a prominent forelock with hair
curls, distinct muzzle incisions, and the placement of four outward-facing
bulls to a cauldron.
In recent years three bronze cauldrons, each preserving Urartian bull
head attachments, and a number of detached Urartian bull head attach-
ments have surfaced, all courtesy of the antiquities market. One of the
cauldrons, together with a tripod stand, is now in the Karlsruhe Museum
(Thimme 1982, 129ff.; van den Berghe and De Meyer 19821983, no 174). It
has three attachments symmetrically placed on the rim. A second, smaller
cauldron is in Japan (Tanabe et al. 1982, pl. 1) and has only two attachments
placed opposite each other. The third, also small, was originally advertized
for sale in London (Apollo, April 1981, 13). Long distance trade eventually
44 Note that the Toprakkale head published by Amandry 1956, 239ff., pls. 24, 26, d, as in
the N.K. collection is now in the Burrell collection, Glasgow, AfO 19 (19591960), 189, fig. 1;
Scottish Arts Review 6, 4, (1958), 27 f.; Amandry 1978, 4: I wish to thank James K. Thomson
of the Burrell collection for answering questions on this matter. Also note that via modern
long distance trade routes the head travelled through at least one other dealer after leaving
N.K.
45 Potentially very important information is not given in van Loon 1989 regarding a
collection of material on loan to the Oriental Institute of the University of Chicago between
19701979, which may have derived from Guschi: why did it take twenty years to publish this
collection, and why werent the owners questioned about their acquisition?
brought it to a dealer in Japan (Tanabe et al., 1982, pl. 2; see also Maass
1978, 11ff., fig. 7a, b; Thimme 1982, 130, 132, fig. 5). This cauldron has four
attachments.
The detached attachments include an example in the Mildenberg collec-
tion (Verdier 1981, no. 18),46 and an example in the Louvre (van den Berghe
and De Meyer 19821983, no. 176). A small example in the British Museum
(Amandry 1956, 260, pl. 32, 3; Herrmann 1966a, 118, n. 10, 128, n. 44) purchased
in 1893 from a vendor, who said it came from Urmia, may also belong to this
group. Solid cast, it is quite small, 3cm. by 2.6 cm; the base is flat and has
eight small, circular depressions arranged in two rows of four. In stylethe
forelock and the hook incision on the muzzleit is Urartian, but if it is an
attachment, now separated from its wing and tail, it is the smallest known.47
The general Near Eastern group is less easily characterized because of the
variety of the head forms and styles and the heterogeneous shapes of the
attachment units, which are either naturalistically rendered, more or less,
as a wing and tail, or are abstracted into a T-shape. Nevertheless, certain
shared features exist: the head and attachment unit are cast together, the
heads are usually embellished with an oval or triangular forelock, and they
were fitted two to a cauldron. There is also a ring cast with the unit to hold
a loop handle, a feature lacking on the Urartian attachments.
Bull head attachments with these characteristics, often recovered on
their cauldrons, have been excavated in Anatolia and North Syria. In Ana-
tolia they occur at Gordion in Phrygia (13 examples), in Tumulus D in
Bayindir-Elmali in Lycia (4 examples),48 in a tumulus burial at Nigde, south
of the Halys River (4 examples), and at the Urartian site of Karmir Blur in
Soviet Armenia (1 pair). In North Syria a pair derives from Tell Rifa"at (found
pressed together with no indication of a cauldron), one from Zincirli, and
one example was purchased years ago in Aleppo.
Of the thirteen Gordion examples at least eightfour pairs on four caul-
drons: three from a tomb, Tumulus MM, and one from the destroyed city
mound (MM 1, 12, 13, TB 1)have bull heads that are very similar to each
other, with a triangular forelock, and a T-shaped abstracted wing and tail
attachment. Two other pairs, on two cauldrons from a rich, probably royal
46 It is not a mate to the Karlsruhe and Japanese examples with three and two attach-
ments.
47 John Curtis kundly furnished the information given here; he also supplied the weight:
66.5 grams.
48 Only one cauldron is published; I was informed that two cauldrons were excavated; see
Acar and Mazur 1989, 150.

tomb, Tumulus W (W 1, 2), and a detached example from Megaron 4 have a

different head form and attachments that are triangular plates (Young et al.
1981, 102ff., 112, 199ff., pls. 50, 59, 87, 88). Kyrieleis and DeVries have noted
the stylistic cluster of the Tumulus MM and TB 1 attachments, and they
have plausibly isolated these as local, Phrygian productions; the Tumulus W
cauldron and the example from Megaron 4 were considered by them to be
imports from North Syria (Muscarella 1989, 340 f.). I would modify this con-
clusion to state possibly from North Syria (infra). Tumulus W is the earliest
Phrygian tumulus excavated at Gordion, dated to ca. 750740 bc (Muscarella
1982, 8). Tumulus MM is, I have argued, datable to ca. 720700 bc (ibid, 9). It
is therefore possible that the Tumulus W imports may have been the inspi-
rational model for the local Phrygian production.
The four examples from Nigde occur on two cauldrons (T.C. Kltr ve
Turizm Bakanligi 1987, 33, figs. 4, 5). Although this area was presumably
outside of Phrygia proper,49 the attachments are apparently identical to
Gordion MM 12 and 13, and they were recovered with manifestly Phrygian
material. The same stylistic fit and juxtaposition to Phrygian material obtain
for the four attachments on two cauldrons from Bayindir-Elmali (Antalya
Museum, June 1989, no. 33 for one example), a site outside of Phrygia.50
From three sites in Anatolia we thus have a significant corpus of bull
attachmentssixteen on eight cauldronsthat may be attributed to Phry-
gia, evidence of an active industry. Furthermore, like the Urartian, the
Phrygian group is well defined.
Concerning the attachments excavated in North Syria, the Tell Rifa"at pair
(Muscarella 1968, 11, fig. 11) differs from the Phrygian group in head shape and
style, and in the presence of a naturalistically rendered wing and tail; here
the heads face into the cauldron. The Zincirli attachment preserves only
the head, which has a round forelock, and the ring for a partially preserved
loop handle; it seems to have faced outward (but I am not sure; pl. IIIab; I
thank Evelyn Klengel for sending me new photographs; see also v. Luschan
1943, 107, pl. 49g). The example purchased in Aleppo (Syria 11, 1930, 366,
fig. 2; Kyrieleis 1977, pl. 29, 4, 5) preserves only the head with a ring. Most
49 Is Nigde in Tabal, and thus is the Phrygian material evidence of Phrygian presence? Or is
the find evidence of gift-exchange or trade with a friendly ally? See Muscarella forthcoming.
50 The finds from these tumulus burials are both extraordinary and unexpected. Aside
from apparent 8th century Phrygian material, the tombs contained griffin protomes (infra)
and one silver and three ivory statuettes of East Greek/western Anatolian style that are
conventionally dated to the second half of the 7th century bc. Were Phrygians buried here?
See n. 49. A preliminary publication of these finds in Acar and Mazur 1989. Pace the quote of
von Bothmer, each piece is not purely Phrygian.
probably it is a local product, as demonstrated by Kyrieleis (ibid., 74, 76); the

combination of its style and area of purchase recommends this conclusion.
I shall return to North Syrian workshops below.
The pair from Karmir Blur,51 recovered on a vessel that disintegrated, is
manifestly not Urartian, in style or construction, and it surely should be
interpreted as an import (Muscarellla 1968, 13; idem 1970, 114; Herrmann
1966a, 129; idem 1966b, 106, 130, n. 165; idem 1975b, 400 [called Phrygian];
Liepmann 1968, 55).52
A rather important find occurred in an unexpected area. At "Ain Gev
in Israel, on the east shore of the Sea of Galilee, in a level dating to the
8th century bc, B. Mazer discovered a fragment of a clay bowl preserving
a schematic bulls head facing the inside (IEJ 14, 1/2, 1964, 31, fig. 11:9,
pl. 14, E, F). At the rear of the head is a molded loop handle which, made
of clay, had no function, but which feature indicates a one-to-one copy of
a bronze model. Thus in Israel, outside of North Syria and Anatolia, bull
attachment vessels, albeit so far known only in clay, were also manufactured.
Winter (1988, 198f.), who mentions this vessel, also mentions two other,
unpublished clay examples from Ai Radanah, also in Israel.
The best known of the unexcavatedunprovenancedNear Eastern
style bull attachments are two on a cauldron in Copenhagen (Amandry
1956, pl. 38). Two more cauldrons with bull head attachments of Near East-
ern form have in recent times appeared on the antiquities markets. One
was advertized for sale in Switzerland and then received as a loan by the
Archologische Sammlung, Archologisches Institut, Universitt Zrich.53
It has four attachments facing outward; on two opposed bulls a winged rap-
tor facing into the cauldron stands on the ring for a loop handle. The second
cauldron first surfaced in Ghent (van den Berghe and De Meyer 19821983,
198, no. 175). Each of the two bull heads face outward, cast with a plain sub-
triangular attachment plate. When the cauldron was traded to Berlin, it was
furnished (how? why?) with a ring tripod stand (Kohlmeyer and Saherwala
1984, fig. 19).
51 In 1968, 7, 13, I incorrectly mentioned only one example, an error I picked up from
several publications. We have no evidence that Urartians manufactured bull attachments

in the Near Eastern manner.
52 Compare the important distinction between the throw-away claims of Kohlmeyer and
Saherwala 1984, 34, that Urartian workshops existed outside of Urartu proper and exported
their productions to Urartu, and the suggestion of Riemschneider 1965, 105f., and Amandry
1978, 4, that Urartians imported cauldrons: both views, of course, given to explain the alleged
presence of sirens in Urartu.
53 Galerie Arete, Antike Bronzen (Zurich, n.d.), no. 21; Isler 1982, 80, pl. 15, 4.
In addition to those in situ on cauldrons, there are a number of unprove-

nanced detached bull head attachments: an example once in the Brummer
collection (Sotheby, November 16, 1964, 66f., no. 160);54 an example in the
Ashmolean Museum (Report to the Visitors 1967, 15, pl. 4a); one in Munich
(Kyrieleis 1977, 74f., pl. 30; Maass 1978, 13f., 17, fig. 9a, b; Herrmann 1984, 23 f.),
which is very close in all features to the Copenhagen cauldron attachments;
one in Barcelona, Spain (Pallej Vilaseca 1979, 202 f., pl. 30); an unpublished
example in Munich mentioned by Thimme (1982, 133) that he says is simi-
lar to a bull head on the cauldron on loan in Zurich;55 a pair, one piece in
the Metropolitan Museum of Art, the other in the N. Schimmel collection
(Muscarella 1968, 7ff., figs. 15); and an example in the Muse du Cinquan-
tenaire in Brussels (Amandry 1969, 800; here pl. IVa). Then there are two (or
four?) unique attachments,56 whose bull heads have a thick band between
the horns, with an elaborate and neatly incised feather and herring-bone
pattern on the wings and tail. They were exhibited in Paris and Ghent (Hotel
Drouot Sales Catalogue, May 22, 1980, no. 383; van den Berghe and De Meyer
19821983, nos. 177, 178).
Let us review what the publishers of these unexcavated and orphaned

attachments, both Urartian and Near Eastern, have offered as contributions
to scholarship. With regard to the Urartian cauldrons, one of the two in
Japan is canonical, that is, it has the expected four attachments. The other
Japanese example and the one in Karlsruhe, however, are, on the basis of
present knowledge, anomalies (Muscarella 1988b, 263, n. 1), for the former
has two, the latter three attachments. Suspicion is necessarily generated
inasmuch as the cauldrons were not excavated (crucial facts ignored by
Thimme 1982 and by Maass 1978, 11f., and Tanabe 1982). However, after exam-
ining the Karlsruhe cauldron, Pieter Meyers reported that he is certain that
the three attachments are in their original position, and he is convinced that
there is no evidence on the cauldron, which he thoroughly examined, for a
fourth attachment. There is no laboratory information about the Japanese
cauldron with two heads.
54 Herrmann 1966a, 157, n. 7 sees a connection with a Greek attachment in the Stathatos
collection. Kyrieleis 1977, 82 compares the latter to 01 A33, which is more acceptable.
55 There is also a small silver bull head in Munich, Maass 1978, 13, fig. 8a, b. Is it a protome
or a cauldron attachment?
56 At first it would seem that only two examples exist, but no. 177 has a broken tail that is
not present on one of the sales catalogue examples. This suggests that there are two pairs
subject to further information.
Vessels with three bull heads exist. A terracotta vessel with three bull head
protomes excavated at Karmir Blur (Piotrovskii 1967, pl. 26) and probably to
be dated to the second half of the 7th century bc may be of local manufac-
ture. But here the heads are placed below the vessels rim, and they lack a
wing and tail.57 There also exists an unexcavated vessel with three bull head
protomes, lacking a wing and tail, placed around its rim (Kellner 1976, pl. 2,
no. 92; Kroll 1979, 1980, 34, fig. 25; van den Berghe and De Meyer 19821983,
no. 194). In all publications it is asserted that this vessel is Urartian; Kellner
even claimed that it came from Patnos! (But he did not say he would return
it there.) But why is it Urartian? And what is its date?58
May we invoke one possible Urartian (Karmir Blur) and one unexcavated
vessel of unknown provenance, both with three protomes, not attachments,
to demonstrate that a stray bronze cauldron in Karlsruhe is normative?59
And what of the cauldron with two attachments? May we objectively reject
Meyers conclusions, which are at least disinterested? Answers do not force
themselves upon us, and one may cite the Karlsruhe and Japanese evidence
only parenthetically as a source of evidence concerning Urartian practice.60
Turning to the unexcavated non-Urartian group, we encounter a problem
concerned with cultural attribution, a problem we shall experience yet again
when discussing the excavated finds from the Greek sphere. It is appropriate
to begin with the Copenhagen cauldron because it plays a prominent role in
the literature. Almost consistently, the cauldron has been assumed by schol-
ars to have been excavated at Cumae in Italy, an assumption that sometimes
produces skewed historical conclusions (Boardman 1980, 169; Coldstream
1977, 206).61 In fact, the cauldron was purchased in the international port of
Naples in 1900. Consequently, it has no known provenance.
The cauldron has been assigned to a number of cultural areas within the
Near East (an indication of the subjective state of art historical analysis)
57 See also a fragmentary vessel that preserves a bull head protome in the Van Museum
(unexcavated), The Anatolian Civilizations I, Istanbul, 1983, 289, A 776.

58 St. Kroll has informed me that he considers this piece not Urartian, but generally Iron
Age.
59 One could assume that the 7th century Karmir Blur terracotta vessel is either an import
or a local copy of a foreign vessel: but I know of no such vessel with these protomes; see n. 51.
60 I am aware that this is not a world-shaking problem: but it is one that focuses on an
ever-present problem in our discipline, that of the methodology employed to gain cultural
information, especially with reference to using unexcavated material in that quest.
61 Coldstream suggests that the cauldron was sent to Greeks in Italy by a Phrygian royal
family; see infra and Muscarella 1989. The fact that the cauldron was purchased in Naples, a
port for many foreign ships, cannot signify it was found in Italy.
but one may emphatically state that wherever it was manufactured, its
workshop will not have been in Urartu (as claimed by Thimme 1982, 133
who cites as Urartian parallels unexcavated Near Eastern style bull attach-
ments; compare Amandry 1956, 243; Herrmann 1966a, 122; Akurgal 1968,
48reversing an earlier attribution to Urartu; Liepmann 1968, 54 f., n. 37;
Calmeyer 1973, 130; and Maass 1978, 14f.: all of whom recognized its non-
Urartian background). Van Loon (1966, 106) inexplicably attributed the caul-
dron to Cyprus; others to Phrygia (Young in DeVries 1981, 219, n. 1; Coldstream
1983b, 206) or to North Syria (Herrmann 1966a, 122, 128; Kyrieleis 1977, 74).
Still others have recorded their hesitancy to place it within a specifically
known cultural center, North Syria for example (Muscarella 1968, 14 f.; Maass
1978, 14f.; Calmeyer 1973, 130).
Stylistic criteria recorded above indicate that Phrygia must be removed
from consideration, which leaves North Syria or another nearby center. I
suggest that we continue to employ the non-specific general term Near
Eastern (Muscarella 1968, 14f.; see also Winter 1988, 198) to identify those
attachments that are clearly neither Urartian nor Phrygian, nor manifestly
North Syrian, in which area there was regional diversity (for North Syrian
style and diversity see Winter 1988).
We may posit at least three North Syrian workshops, not solely because
of the find spots (two sites, plus the Aleppo purchase), but because of the
differences in the head form among the attachments. They give us some idea
of the variety of related styles/forms being produced in cauldron workshops
within North Syria. Variation in form and execution within a stylistic unity
is a characteristic of other metal workshops and also of ivory carving from
that area (Winter 1988, 194ff.).62
A concern with origin also must pay attention to the accident of archae-
ological recovery, in this case to the lack of excavated material that might
indicate the existence of other manufacturing centers outside of Phrygia
and North Syria, in adjacent south central Anatolia for example. Indeed,
the finds from "Ain Gev and Ai Radanah mentioned above are telling argu-
ments against casually offering attributions to the presently known centers.
But while compelled to admit that unrecognized workshop centers prob-
ably existed, we are constrained from hypothesizing where they were and
purport to know their products (Tabal is becoming a popular guess-area).
62 See also Winter 1981, 101 ff.; Winter 1983, 179 ff. Note a bronze wing and tail attachment
from Hama (Riis and Buhl 1990, 120 ff., no. 390). It has a ring at the rear, but if there was
originally a protome, it is missing.
In short, the Copenhagen cauldron mayor may not have been made in a
North Syrian workshop; it is a Near Eastern cauldron. The same conclusion,
I believe, must obtain also for the Karmir Blur attachments and those from
Tumulus Wand Megaron 4 at Gordion. The Tumulus W examples have
triangular decoration above and below the forelocks, the same decoration
found on the sirens; but the head-wing-tail units are claimed not to be cast in
one piece, apparently a unique feature for a non-Urartian bull attachment.
If they were made in North Syria, it will not have been in one of the three
workshops mentioned above.
It follows, accordingly, that Herrmanns North Syrian Cumae Gruppe
(1966a, 128f.; 1984, 23 f.) is misleading, not merely because of the improper
use of Cumae, but because the alleged group has no demonstrated stylis-
tic unity (on this v. also Maass 1978, 14ff.; Rolley 1984, 282). Included in
the group are sirens of disparate forms, and, I suggest, attachments made
in several distinct cultural centers reflected in the differences in styles
Greek, North Syrian, Phrygian, and Near Eastern. The same mixture of styles
compromises Herrmanns weitere spthethische Werksttten (1966a, 128 f.;
1984, 24): Greek, Cypriot, Phrygian, Near Eastern. What is rather extraordi-
nary is that Herrmann in 1966 did not cite the excavated bull attachments
from North Syria in his discussions of the cauldrons and their attributions to
that area; they were first mentioned by him in 1980 (576; see also 1984, 23 f.,
where he assigns the Aleppo head to the Curnae Gruppe, the Zincirli and
Tell Rifa"at heads to a second and third North Syrian workshop).
The anonymous writer in the sales catalogue advertizing the Near Eastern
cauldron later loaned to the Zurich University collection (n. 53) identified
it as Urartu, um 700 v. Chr., but then proceeded to associate it with North
Syrian attachments, the latter opinion accepted by Herrmann (1984, 26) and
tentatively by the author (1988b, 263, n. 1). In style, the bull heads could indi-
cate a North Syrian background,63 but the placement of four non-Urartian
heads, two embellished with raptors, is unique and not archaeologically
documented anywhere. It follows that we rather should avoid attributing
this orphaned cauldron, whose homeland may never be known.
The same conclusion is also in force for the equally unique Berlin caul-
dron with its two bull attachments (pace van den Berghe and De Meyer
19821983, 108, and Kohlmeyer and Saherwala 1984, 34 f., who infelicitously
63 For example, compare it to a stone bull head in Woolley and Barnett 1952, pl. B47:
muzzle swelling, eye form, round forelock, and head proportion. Herrmann 1966a, 122, n. 22,
cites this stone bull head as a parallel to Olympia A31, which it is not.
label it Urartian; cf. Herrmann 1984, 26, n. 59a, who rightly rejects this
attribution). There are no close parallels for the stylized heads and they are
difficult to attribute.
Both the Brummer and Ashmolean detached winged bull heads are prob-
ably Near Eastern (not Urartian, pace Kyrieleis 1977, 73 on the latter head).64
Here too we may posit that they were made in North Syria or somewhere in
Anatolia. The Brummer head seems to be Near Eastern in form but is unlike
any of the others known to me.65
The Munich attachment is indeed very similar in almost all features to
those on the Copenhagen cauldron, as Herrmann, Kyrieleis and Maass have
recognized (the latter scholar alone not accepting a North Syrian back-
ground). The only obvious difference is the wing-tail attachment unit: natu-
ralistic wing-like outline and herringbone pattern on the Copenhagen fig-
ures, smooth and plain on the Munich one. Although the attachment is
unexcavated, the three above mentioned scholars (also Kilian-Dirlmeier
1985, 248) accept an Olympia provenance. This claim, like the one that the
Copenhagen cauldron was excavated in Cumae, complicates an already con-
fused situation; in both instances we do not know if they were plundered in
Greece, Italy, or the Near East.
The Barcelona attachment is probably, as Pallej Vilaseca suggested, an
orientalizing Greek product (cf. Herrmann 1984, 24); it was purchased for a
private collection, and one cannot state that it was archaeologically recov-
ered in Spain. Perhaps also reflecting a Greek, perhaps East Greek or western
Anatolian, background are the two/four unique bull attachments exhibited
in Paris and Ghent. Wherever their origin, however, stylistic determination
precludes an Urartian attribution, as published (v. Kyrieleis 1977, 79, 82 for a
discussion of the Greek background of the Stirnband; for the wing structure
see Akurgal 1961, fig. 125, 153, 154, 210). The example in Brussels (pl. IVa) has
a plain wing and tail, small upright horns, and a flat elongated head ending
in a thick muzzle; it is not Near Eastern.66 Although it is by no means certain
that the Metropolitan Museum-Schimmel pair derived from Iran (even if it
was purchased there, Muscarella 1968, 7), stylistic analysis does not argue
against such a background (ibid, 17f.). The pair remains unique and may be
late manifestation, 6th century bc.
64 The style of the head and its casting technique are foreign to Urartian bull head
attachments.
65 The closest parallel I know is a goat (?) attachment in the British Museum, BMQ 37, 3/4
(19721973) 122 f., pl. 50, b; it is unprovenanced and unique.

66 I thank D. Homs-Fredericq for sending me a photo of this piece. The Brussels example
is also mentioned by Goldman 1961, who accepts it as coming from Luristan!

Finally, a brief comment is warranted concerning some still unpublished

attachments. Among a group of 98 bronzes purchased by the University
Museum of the University of Pennsylvania are three winged bull head
attachments (Muscarella 1968, 18, n. 28; DeVries 1981, 220).67 The whole col-
lection is considered in the Museum records to consist of Urartian material.
The heads are different in size and only one (inv. 67-39-3) has a ring for a loop
handle; thus they derive from three different cauldrons.68 The head with the
ring and another (inv. 67-39-4) are close enough in form and style to have
come from the same workshop. All three have the head-wing-tail attach-
ment cast in one piece, and all have a rounded forelockcharacteristics of
Near Eastern but not Urartian bull attachments. There can be no archaeo-
logical claim that the purchased collection is a bona fide hoard (where are
the mates to the three heads?), although most of the other objects seem to
be Urartian artifacts.69 Speculation does not help resolve the question how
the attachments got into the collection.70 (See Strms paper, infra, 54, n. 29,
and supra, 18, n. 9.)
Bull head attachments excavated at sites in the Greek sphere are just about
equal in number to those excavated collectively in the Near East, Olympia,
12 (Herrmann 1966a, 114ff.; 1984, 22ff., A24a, A34a);71 Delphi, 8 plus (Per-
drizet 1908, nos. 327332; Syria 35, 1958, pl. 6c, d; Rolley 1984, 280 mentions
several unpublished oriental examples); possibly one from Athens (Herr-
mann 1966a, 129);72 Argos, two (Waldstein 1905, pl. 75: 23, 25);73 Amyclae, one
67 See zgen 1984, 91 ff., where some of these objects are published; in n. 4 on 119, zgen
refers to his dissertation on some of the material in the collection. Note that neither in the
1984 article nor in the dissertation are bull heads mentioned. The bull heads are on exhibition
in the University Museum.
68 One of the heads, inv. 67-35-3 is claimed in the Museum records to fit onto three holes
preserved on a cauldron fragment in the collection.

69 The collection was purchased from a dealer acting for another: how many other
handsnot to mention those of the plundererswere involved is unknown. Note that

Amandry 1969, 800, refers to three bull attachments he saw on the antiquities market: could
they have been the three mentioned here?
70 See also an unexcavated bronze bucket in Japan (Tanabe et al. 1982, pl. 3) with two
winged animal head attachments (bulls? but no horns are visible). Objects like this one and
that mentioned in note 43 are known only from modern plunder, not from excavations.
71 Note that from Herrmanns 1966a and 1984 total of 14 examples I believe we should
remove A28 and A34 because they are not clearly attachments, as Herrmann 1966a, 117f., 121,
126, himself realizes and which there he thinks might be Greek works.
72 Is this actually a cauldron attachment? See Amandry 1956, 245, n. 17.
73 There is a winged attachment on pl. 123, no. 2204, that deserves a better publication: is
it a lions head, as claimed?

(Muscarella 1968, 9, figs. 6, 7); Delos, one (Rolley 1973, 513 ff., figs. 20, 21);
Lindos, two (Blinkenberg and Kinch 1931, nos. 704, 706), and Samos, 11 or
more (Jantzen 1972, pls. 76, 77; Herrmann 1966a, pl. 52, 2; AM 74, 1959, Beil.
71: 1; Kyrieleis 1977, 83f., pls. 3537; Isler 1978, 77, no. 16, pl. 38).74 These
figures add up to 38 bull attachments, plus the unpublished Delphi examples
and perhaps others from Samos.75 Only one example, Olympia A31, was
recovered in situ on a cauldron.
Cyprus has an isolated example from Idalion, two on a cauldron probably
from Kourion, and two units of three each on a cauldron from Salamis
(Catling 1964, 154f., pl. 21, c; Karageorghis 1973, 108 ff., figs. 2528, pl. 246).
With the exception of Herrmann 1966a (infra),76 all scholars agree that
Greek workshops produced cauldrons with bull attachments, copied, of
course, from the oriental imports and paralleling the history of siren caul-
drons. But the literature reflects much disagreement and confusion once
attempts are made to isolate the imports from the locally made pieces. This
problem is well formulated by Kyrieleis (1977, 80) as Die Frage Orient oder
Griechenland? This question leads naturally to another, one that also fol-
lows from the previous discussion of Near Eastern workshops: whence in
the Near East did the imports depart?
In 1968 (15, n. 70), in reaction to Herrmanns claims (1966a, 124, n. 23, 126;
1966b, 136) that not only were all the bull attachments from Olympia oriental
imports, but that in fact there is no evidence that the Greeks copied them,
I countered that indeed there were only about four attachments from the
Greek sphere that were imports. I cited the example from Amyclae, Samos
BB 740, Olympia A24, and no. 23 from Argos; and I cited a few about which
I was uncertain, Delphi no. 329, Argos no. 25, and the Cypriot example from
Idalion. It was also noted, and those scholars who have paid attention to
style and technique of manufacture agree, that not a single example of a bull
attachment recovered in the Greek sphere relates to the distinct Urartian
group. Since 1968 more discussion has ensued, and a review of my list of
imports and those of others follows.
First of all, it should be appreciated that using traditional stylistic anal-
yses does not produce firm conclusions. This is because, as recognized by
Herrmann (1966a, 121ff., 124) and Kyrieleis (1977, 71 f., 85 f.), it is far easier to
74 The bull protome mentioned by Kyrieleis on 81, n. 59 is apparently not an attachment.

75 This figure includes bull attachments with single heads and a few with double heads
that have been discussed by the authors noted here.
76 But see 1984, 23, n. 38, where he accepts Delphi no. 373 as Greek made; he does not
comment on his earlier view.

perceive and isolate distinctive traits in the style and execution of human
headseyes, mouths, hair, facial castthan of bull heads. While to my
mind (and to Kyrieleis) the majority of bull attachments in Greece are of
Greek manufacture, a certain number seem not to be. But which ones? For in
the course of studying the evidence and attempting objective attributions,
I realized that I changed my mind several times, which indecision precisely
indicates the problem.77
The oriental (Near Eastern) origin of the Amyclae attachment is accepted
by Herrmann, Kyrieleis and Maas. Kyrieleis (1977, 87) also agrees that Samos
BB740 is an import, but to him from Assyria, an attribution that has no stylis-
tic underpinning. I suggest it fits into the same Near Eastern background
that produced the Copenhagen cauldron and the Munich example (see also
Calmeyer 1973, 130). The presence and position of the volute on the wings,
however, allows us to make a more specific claim, namely that the Samos
attachment points to a North Syrian workshop. I am still inclined to think
that Olympia A24 and a recently discovered similar example, A24a (Herr-
mann 1984, 22, pl. 6: 1, 2)78 are Near Eastern; the form and position of the ears
and horns are very like those on the Tell Rifa"at heads, although the Olympia
heads are more attenuated. Kyrieleis (1977, 86, n. 93) is uncertain whether
A24 is Greek or oriental, and he calls it a Zweifelsfall, noting the lack of a
forelock.
I no longer think that Argos no. 23 is oriental, mainly because of the dis-
cussion of the Stirnband/Stirnrosette by Kyrieleis (1977, 79, 81 f., 84). Kyrieleis
(1977, 87) also shares my uncertainty whether Delphi no. 329 and Argos
no. 25 are Greek or oriental. Both pieces have the characteristics of the
orientalische Typus, including the forelock, but, echoing my own feel-
ings, Kyrieleis wonders whether es sich um etwas provinzielle griechlis-
che Nachahmungen orientalischer Vorbilder handelt. The execution of the
heads and wing shapes of both examples do readily fit into a Near Eastern
background.
After reexamining the Idalion head (not autopsy), I have no doubts but
that it is a local work (pace Herrmann 1966a, 123, n. 26),79 as are all the other
Cypriot examples cited (for oriental influence on the local bull attachment
production see Herrmann 1966a, 123f.; Liepmann 1968, 54 f.).
77 See also Herrmann 1984, 23, n. 38 and 25, n. 50.

78 They are not a pair but probably derive, as Herrmann claims, from the same workshop.
79 Catling 1964, 154 f., sees parallels with the Urartian attachments, but then refers to North
Syrian trade with Cyprus. Liepmann 1968, 55, sees it as being close to Herrmanns Cumae
Group.
Of the twelve examples recovered at Olympia, three, aside from A24 and
A24a, warrant further comment: A31, A34, A34a. Herrmann (1966a, 120 f.,
124; 1984, 23, n. 38), noting among other features the dewlap folds, strongly
argued that Olympia A31 was not a Greek work. Among those who discussed
the piece after 1966, only Maass (1978, 17) agreed it was oriental; Rolley
(1973,515), and later Kyrieleis (1977, 76f.), with extensive analysis, argued that
it was made in Greece.80 This is the position I took in 1968 (14, n. 20) and still
maintain, all the more so after Kyrieleis cogent analysis.
Likewise, I maintain that Olympia A34 is a Greek work. Originally per-
ceived as Greek by Herrmann (1966a, 121, 125f.),81 he reversed himself in
1984 (25). In the 1984 article Herrmann published another bull attachment
from Olympia, one similar to A34, that he labelled A34a, and which he also
attributed to the Orient. Maass (1978, 14) considered A34 to be securely
Greek, while Kyrieleis (1977, 79, 86) accepted A34a (he did not mention A34)
as oriental. To Kyrieleis, A34a and the unexcavated attachment in Munich
are the only certain (!) oriental examples in the Olympia repertory.82 My eyes
cannot find stylistic parallels that would securely place A34 and A34a in the
Near East (see also n. 71).
Delphi no. 373 is surely Greek, as noted by all the above cited schol-
ars, even now Herrmann (1984, 23, n. 38), who rejected his 1966 oriental
attribution.83 Delphi 327 and 328 (pls. IVbc, Vab) are considered to be
North Syrian by Herrmann in all his publications, and also by Kyrieleis (1977,
87),84 who sees a very close relationship to the Zincirli and Aleppo heads. I
doubted this conclusion in 1968 (14, n. 20), and I do not see the relation-
ship suggested by Kyrieleis: cf. pls. IIIab, IVbc, Vab. C. Rolley informs,
me that no. 327 preserves a small section of what must have been a short
80 No one has mentioned the claim by Herrmann 1966a, 7, 25, 142 that the cauldron to
which the attachment was joined (they were found separately) has placements for five bull
attachmentsa feature not hitherto documented, neither in the Near East nor the Greek
sphere. Herrmann (142) denies that bull cauldrons ever appear with protomescould this
cauldron be an exception?
81 Kyrie1eis 1977, 86, claimed that Herrmann recognized no Olympia attachments as
Greek-made; see also here n. 71.

82 In reality, to Kyrieleis, only one example, A34a, and possibly A24, are oriental imports
at Olympia. While Herrmann thinks A34 may not be an attachment (see n. 71), he makes no
such claim for A34a, which seems to be an attachment.
83 Kyrieleis 1977, 84 f., sees it as East Greek; he is uncertain whether it belonged to a
cauldron.
84 There is no indication that either Herrmann or Kyrieleis have actually examined the
Delphi attachments. Delphi nos. 330 and 331 remain unpublished and are not discussed here.
I wish to thank Claude Rolley for sending me photographs of Delphi Nos. 327 and 328.
T-shaped attachment plate and evidence that there was once a ring at the
back of the head; no. 328 preserves only the head. No forelock is evident in
the photographs, although no. 327 is badly damaged. I cannot relate these
heads to the Near East, but as they are both poorly preserved, I will leave the
issue of origin open.
The Delos example, considered to be North Syrian by Herrmann (1975a,
307), is to my mind certainly Greek (also Rolley 1973, 513 ff., no. 18, figs. 20, 21).
An example from Lindos (Blinkenberg 1931, no. 706) remains unpublished.
It was described as having a wing and tail like the Lindos siren attachment
no. 705 (Herrmanns no. 54, supra); not being published, one cannot discuss
it. Lindos no. 704, a double protome attachment, is clearly not Near Eastern
(cf. Herrmann 1966a, 125, 129, pls. 44, 54).
We turn now to Samos where, as may be expected, there is also contro-
versy concerning attribution: local or import? I consider three of the eleven
bull attachments from Samos to be Near Eastern imports, BB740 (supra),
B1266, and, highly probably, B438 (Jantzen 1972, pls. 76, 77). Herrmann (1980,
577f.), Kyrieleis (1977, 87) and Kilian-Dirlmeier (1985, 250) agree that B1266
is an import from Phrygia (v. also Muscarella 1989, 341). All also agree that
B348 is North Syrian; surely, if not a close local copy of an oriental model,
it must be Near Eastern, either from North Syria or another close-by work-
shop. Kyrieleis (1977, 80, 83f.) argues that B1482 (Jantzen 1972, pl. 77), B161,
B171, B172, and B303 are all Greek or East Greek, a conclusion I accept. B51
(Jantzen 1972, pl. 7) is a fragment lacking wing and tail (and face?), yet both
Herrmann (1975a, 307) and Kyrieleis (1977, 87, n. 101) see it as North Syrian,
Kilian-Dirlmeier (1985, 252) as oriental; I am not convinced it is oriental.
Calmeyer (1973, 130) accepts B348, B51, B1482, and B1266 as oriental, to him
from east Anatolia (but not Urartu). Samos then has, I suggest, at least one
example from Phrygia (B1266), one from North Syria (BB740) and one (or
more) general Near Eastern (B348)representing at least three imported
cauldrons.
Summarizing the Greek-oriental discussion, it is concluded in this review
that of the bull attachments recovered in the Greek sphere, no more than
six, representing six cauldrons, appear to be imports from the Near East:
Amyclae, Olympia A24, A24a, and Samos BB740, B1266, B348. Five others are
of uncertain origin and require more investigation and better publication:
Argos no. 25, Delphi nos. 327, 328, and 329, Samos B51.
Kyrieleis (1977) has nine definite imports in his list, representing nine
cauldrons, Amyclae, Olympia A34a, and the unexcavated bull attachment
in Munich; the same attachments from Samos as on my list plus B51; Delphi
nos. 327 and 328. He is uncertain about Olympia A24, Argos no. 25, and
Delphi no. 329 (the last two are also on my uncertain list). However, one
reacts to the respective views, what seems almost certain is that Samos has
both imported and locally made bull cauldrons and Amyclae has an import;
Olympia, Delphi and Argos may have both.85 Furthermore, only two imports,
Samos BB740 and BB1266 may be attributed with some certainty to specific
Near Eastern polities, North Syria and Phrygia.
Griffin Protomes
Although the griffin protome is inextricably linked to the history of the siren
cauldrons (v. supra), it remains a major and vertiginous problem (along with
the lion protome).86 I am speaking about the origin, oriental or Greek again,
of the actual protome, not about the griffin motif or concept, whose origin is
manifestly Near Eastern (but not Urartian! Muscarella 1962, 319 ff.; 1981a, 50 f.;
Herrmann 1979, 137ff.). The question is whether oriental cauldrons arrived
in Greece with hammered griffin protomes and sirens as an ensemble, or
wether the protomes were added to the siren cauldrons in Greece to satisfy
(unknown) local needs.87 I will not discuss this problem here because I have
little to add (v. the discussion supra on the Munich siren cauldron), except
to note that Bensons (1960, 65) and Herrmanns (1966a, 137 f.) claim that no
cauldrons with siren attachments together with griffin protomes have been
recovered in the Near Eastthis ensemble occurring only in the Weststill
obtains.88
Cast griffin protomes do of course occur in the Near East, and since
1979 a few more have surfaced; none are earlier than the earliest examples
recovered in Greece, and none are considered to be of local manufacture.
In addition to the already recorded cast examples from Susa, Miletus, and
Ephesus (Herrmann 1979, nos. 217, 131, 232, 242, 304, 349, 350, 143, 177 and 400
85 Olympia seems to have a majority of Greek bull head attachments, while Delphi may
possibly have a majority of oriental onesjust the reversal of the siren proportions; but this
will only be resolved by full publication.
86 To give but one example of the problem of attributing lions to a source, in 1981a, 52. I
stated that an oriental lion protome from Olympia, B4999, was North Syrian: based on a photo
published in ILN. The protome is now published in detail with good photos in Herrmann 1981,
72 ff., pls. 47. I now am less sure about its North Syrian origin as it is not typical for that area.
While I do not exclude that area, I cannot assert it.
87 Muhly 1988, 338, confuses the origin issue between hammered and cast griffin pro-
tomes; cast protomes are accepted by all as Greek, it is the hammered ones that are contested,
Greek-made or oriental import.
88 See n. 20.
[?]), and examples in the Ankara and Izmir Museums89 (ibid., nos. 215, 245
and 302), we now have more excavated examples from Ephesus (Herrmann
1984, 32; Bammer 1984, 201, figs. 63, 64), from Samos (unpublished), two
from Ikiztepe near Ushak in western Turkey (unpublished), and two from
the extraordinary find from Bayindir-Elmali (Antalya Museum June 1989, 34,
187, nos. 29, 30).90 These last two are described as being cast; their date of
manufacture remains to be investigated.91
Herrmann (1984, 26ff., 32f.) has listed newly recovered hammered and
cast griffin protomes from the Greek sphere, from Olympia, Samos,
and Dodona, and some newly surfaced unexcavated examples. Hammered
and cast examples (unpublished) exist at Isthmia. There are now over 450
griffin protomes known to exist (cf. Muscarella 1981a, 47).
Chronology
Herrmann (1966a, 3f., 148) sees the beginning of siren cauldron production
in the Near East sometime in the third quarter of the 8th century bc; he
believes that export to Greece also began at that time (see Strms paper,
infra, 47f.). Chronological evidence available for close review from a Near
Eastern center exists solely in Tumulus MM at Gordion. There is continuing
debate about the precise time of the tombs deposition. I have dated tumulus
MM to a time before the destruction of the site in ca. 696 bc, to ca. 720700 bc
(1982, 9f.), while other scholars (ibid., 9) date the deposition either imme-
diately after the ca. 696bc destruction, or to a time before a later possible
destruction date, ca. 675 bc.92 Herrmann (1966a, 86 f., 129) also supports the
earlier dating of Tumulus MM to arrive at his conclusions.
The same chronology for the same reasons, namely their presence in
Tumulus MM, obtains for the Phrygian bull attachments (Herrmann 1966a,
129f.). But Tumulus W, ca. 750740bc (supra), contained, as already noted,
89 Hanfmann 1957, 241, says the Izmir example comes from Labranda.
90 No cauldron is mentioned, or evidence of other fragments.
91 They derive from Tumulus C (the bull cauldrons came from Tumulus D). The griffin
heads look similar to the hammered heads in Herrmann 1979, G25 and G32, dated by him to
the early decades of the 7th century bc.
92 Without discussing the archaeological remains, Snodgrass 1971, 350 accepts a ca. 675bc
date for the destruction of Gordion, and he dates Tumuhls MM to ca. 685/680 bc. If this late
chronology were accurate, it would affect our perception and knowledge not only of Phrygian
history but also of the nature of Greek-Phrygian relations. Thus, Snodgrass sees Phrygians
playing a minor role in Greek orientalizing because of his low chronology; he does not discuss
the Midas gift.
two bull cauldrons that seem not to be Phrygian, but North Syrian or general
Near Eastern. Whatever their cultural origin, they are the earliest dated bull
cauldrons known to exist, and it seems from the Gordion sequence that bull
cauldrons were manufactured earlier than siren cauldrons.
How long production of bull cauldrons continued in Phrygia is unclear.
We may plausibly assume that the presence of the three examples in the
debris of the destroyed citadel indicates production down to the early 7th
century bc, if they were not curated heirlooms. Precise dating of the Bayin-
dir-Elmali tumulus, the deposition of which could date to a time in the 7th
century, perhaps even to the second half, will have some bearing on this
issue (see nn. 50, 91). Only a general late 8thearly 7th century chronology
appears to exist for the Urartian bull attachments, and one of the alleged
key chronological supports has been eliminated. Evidence for a possible
late 8thearly 7th century date is the poorly documented find from Alishar,
where a bell inscribed with the name of Argishti was apparently recovered;
but it is uncertain whether this is the first (ca. 780750bc) or the second
(ca. 712685 bc) king with that name (Muscarella 1978, 64, fig. 2). The Alt-
intepe tomb was for some time thought to date to ca. 740732 bc because
of the presence of an inscription allegedly of Urikki of Cilicia. But it has
been known since Klein 1974 (87ff.) that the name Urikki in fact does not
appear, which eliminates Altintepe as a chronological datum.93 Toprakkale
was founded by Rusa II (ca. 685670bc) not Rusa I (van Loon 1989, 268).94
The tomb is thus early 7th century in datealthough the cauldron could
have been an heirloom. For subjective relative chronologies, or sequence of
manufacture, of the Urartian bull attachments based on style, see the differ-
ent views of Hanfmann 1956, 212; Azarpay 1968, 53; Verdier 1981, 31; Thimme
1982, 133; van Loon 1989, 268.
Of the North Syrian finds, only the examples from Tell Rifa"at may be said
to yield chronological information, although it is not very specific: 9th7th
centuries bc (Winter 1988, 198, 208). But an 8th century date is supported by
the "Ain Gev clay example, which form presupposes the existence of metal
bull cauldrons at this time.
The Tumulus W evidence documents that bull cauldrons were being
manufactured in the Near East in the mid-8th century bc, and other finds
from Gordion, Urartu, Israel, and North Syria indicate a continuous produc-
tion and use to the end of the century and later.
93 Klein 1974 was ignored by DeVries 1981, 220; Brookes 1982, 608; and van Loon 1989, 268.
94 See also Salvini 1988, 130, 135 f.
A review of North Syrian history in the late 8th century bc may yield
information concerning the chronology both of local production and the
period of export, assuming, of course, that North Syria was the source of the
siren cauldrons and of some of the bull cauldrons. Winter (1976, 17 ff.) has
noted that the production of North Syrian ivory and metal artifacts seems
to taper off, or may have ceased, in the late 8th century bc. She links this
situation to the Assyrian destruction of North Syrian cities by Sargon II,
beginning in 720bc with Hamath, followed by Carchemish in 717 bc, Malatya
in 712 bc, and so forth. These destructions must have interfered with local
workshop productions. The date of the destructions suggests the possibility
that the export of the cauldrons to the Greek sphere could have been in force
before the Assyrian onslaught and ceased, or at least declined, sometime
in the penultimate decade of the 8th century bc. This chronology is not
readily confirmed from stratigraphical excavations on the Greek side where
griffin cauldronsbut not those with siren and bull attachmentsare first
depicted in art in the late 8th (presumably) and early 7th centuries bc
(Herrmann 1966a, 1f., figs. 14).
Methodology
Although it arises often in art historical analyses, one infrequently ponders

in print upon a problem intrinsic to the conclusions offered here and else-
where on the Greek-oriental problem: the subjective eyes vs. eyes basis
for identification and attribution. On the basis of perceptions of style it has
been argued that to a considerable degree Greeks copied both siren and bull
cauldrons, the models for which came from foreign cultures in the western
regions of the Near East, Anatolia and North Syria. But the Greek cultural
milieu would have to be perceived as markedly different if, on the one hand,
we were to accept Herrmanns conclusions that the Greeks made relatively
few siren cauldrons and no bull cauldrons,95 and on the other hand that
Urartu was one of the oriental cultures involved both in the initial export
and the eventual cultural stimulus that we call orientalizing. Herrmann
1966a and b and Kyrieleis 1977 are serious contributions to the ongoing
study of orientalizing; neither is a casual engagement with the evidence, but
95 Part of the problem here is that there may have been more manufacturing workshops
in different areas of the Greek sphere than hitherto recognized; see Kyrieleis 1977, 77f., 79f.,
83 ff.
in many instances they arrive at different stylistic, and therefore historical,

conclusions. My own observations have resulted in a basic agreement with
Kyrieleis 1977 against Herrmann 1966a on bull attachment attributions, but
with Herrmann 1966a and b against Kyrieleis 1967 on the issue of the siren
cauldron attribution. However, I call attention to the subjective nature of
the decisions, and recall that both Herrmann and I have changed our minds
in a few instances after years of review and reflection.
The issues involved are not academic. It is not abstract art historical
activity that compels the attempt both to distinguish the import from the
local production, and to identify which cultures in the Near East played
a role in orientalizing Greece and which did not. The correctly defined
import yields evidence of a contact of some form and degree between
Greece and the Orient; its presence in a sanctuary informs us that some local
entity accepted the object. The copy or adaptation indicates influence,
a heightened and more complex consumption of the original import, an
absorption into the local culture. Yet, while the ancient Greeks knew the
differences between an import and their own work, whether they were
always aware of the specific cultural sources and the original ideological
message of the imported works eludes us (see n. 105 and infra).
One might think that metal analyses of trace elements would contribute
independent and objective evidence to solve questions of origin, but a lay-
man in this respect, I find the various conclusions presented to be confus-
ing and contradictory and as subjective as are art historical conclusions.
Thus Hughes et al. (1981, 144) claim that their analyses of Urartian bronzes
indicated a general resonance of agreement between them and those of
Assyrian background, while Gale et al. (1983, 50) claim that their analyses
of Urartian bronzes show correlation with oriental bronzes (non-Assyrian)
at Delphi. Steinbergs (1981, 288) analyses show that trace elements of Phry-
gian bull attachments (MM 1) are similar to those from the sirens recovered
in Italy, but zinc content (ibid., 289) is similar to that of other Phrygian
bronzes as well as some Urartian bronzes. Winter (1988, 204) interprets
Steinbergs work to indicate that the cauldrons analyzed were not made in
Urartu and Phrygia, but probably in North Syria. And Romano and Pigott
(1983, 129) claim a consistency in trace elements among oriental sirens from
Italy, Greece, and Gordion and suggest they were all made in the same locale.
The last two conclusions fit neatly into the conclusions of this paper,
but it would be premature to accept them unexamined. In the analyses of
sirens the tin content differs considerably from piece to piece (Romano and
Pigott 1983, 128; Steinberg 1981, 287) even on the same cauldron (Gordion);
one siren from Gordion (MM 2) has 25% tin, more than twice that of other
Gordion sirens. Furthermore, the percentage of lead varies considerably

among the sirens from Italy, Greece, and Gordion, the latter having only
trace amounts.
Then there are the analyses and conclusions of Filippakis et al. (1983),
where it is claimed that one is able to distinguish Greek from oriental and
orientalizing bronzes of the 8th century bc by the percentage of tin present.
Their conclusion is that tin occurs only after ca. 750bc and that this tin had
been introduced by oriental immigrants (see also Magou et al. 1986, 122,
133). But since these authors are constrained to admit that tin does exist
earlier in typical Greek bronzes, they conclude that in these cases the tin
was introduced by the remelting of Mycenaean bronzes (see also Magou et
al. 1986, 122). Here a patently undemonstrable explanation is employed to
support an alleged scientific observation, a method not proper in scientific
analysis (v. Muhly 1988, 338), and this explanation compromises the report
and its conclusions.
Steinberg (1981), Winter (1988), Muhly (19801983, 357 f.) and Craddock
(1988) have forcibly reminded us of some of the problems involved in the
interpretation of metal analysis. There are events in the process of the
procurement of ores and in the casting procedures that we cannot chart:
the possible use of remelted bronzes, which may derive from one or more
sources, and the impossibility of recognizing this activity; the uncertainty
of recognizing which trace elements derive from the tin, which from the
copper; the possibility of a single mine source having different elements in
different parts of the vein; the segregation of elements in areas tested or not
tested; and so forth.
At least for the present, metal analyses do not supply answers to the ques-
tions archaeologists have been asking about the originplace of manu-
factureof cauldrons with attachments.
Function
Although we are not able to infer all the implications connected with the
presence of the oriental cauldrons in the West, their ultimate function
seems relatively easy to ascertain. Evidence from many excavations in the
Greek sphere indicates a pattern: the cauldrons occur only in sanctuaries
of various deities, male and female (see Strms paper, infra, 52 ff., fig. 5),
and they served as votive objects deposited as dedications (for a different
interpretation, banqueting, see Strms paper, infra, 55 f.). It is fundamental
to recall here that the dedication of cauldrons to deities at their sanctuaries
was a Greek custom in existence before the time when the first oriental
examples arrived (Coldstream 1977, 334ff.), and continued in later periods
(Herodotus IV. 152).
In the Near East the cauldrons apparently had functions that were similar
to those known in the West. At Gordion, siren and bull cauldrons were
deposited in tombs, and bull cauldrons were also recovered on the citadel.
Recent cogent research on the contents of Tumulus MM (Simpson 1986, and
1990, 6987; v. also Moorey 1980, 195) suggests that most likely the cauldrons,
and accompanying vessels, were used in a funeral banquet for the deceased
before they were placed in the tomb (see Strms paper, infra, 55). At first
examination, it may seem that a tomb deposition precludes a dedicatory
function, which is Herrmanns (1966a, 149) meaning when he speaks of a
sepulkrale Funktion in Phrygia as opposed to use as Kultgerte in Greece.
This conclusion, however, may be too restrictive. In Tumulus MM, and in
the burials at Nigde and Bayindir-Elmali, after use in the banquet, which
itself may have been a religious ceremony, the cauldrons may have been
deposited either as utensils for personal use by the deceased, or as votives to
be presented as dedications (gift exchange) to gods encountered in the new
life.96 In other words, albeit not manifest, a dedicatory function for cauldrons
may have existed in Phrygia. This interpretation may also obtain for the
cauldrons deposited in tombs in Cyprus and Italy.97
As for Urartu, the cauldron with bull attachments from Toprakkale was
recovered in the ruins of the Haldi temple (Barnett 1950, 1 ff.), which locus
indicates a dedicatory function for cauldrons there. Secondly, at Kayalidere
a stone base facing the temple has three triangularly arranged hoof-shaped
depressions for a tripod (Burney 1966, 72, pl. 8b). Furthermore, we recall that
the inventory list of Sargon IIs booty from Musasir mentions cauldrons on
tripods among the dedications in the Haldi temple (van Loon 1966, 85).98
All the other excavated Urartian cauldrons derive from tombs, certainly
at Altintepe, and most probably also at the sites of Guschi and Alishar. In
Urartu, then, we have evidence of bull cauldrons placed both in sanctuaries
and in tombs.
96 Along with the bag of fibulae placed behind the deceaseds coffin, Muscarella 1967b, 1,
53, 51 f., n. 19. For fibulae as votives see Muscarella 1967a, 85 f., and Muscarella 1986, 199.
97 The fact that animal bones were found in one of the Vetulonia cauldrons does not, of
course, preclude a dedicatory function.

98 No attachments are mentioned, and plain cauldrons were depicted on the well-known
relief illustrating the sack of the Haldi temple (van Loon 1966, 44), but the fact remains that
cauldrons were placed in the temple complex.
The examples from North Syria, Zincirli and Tell Rifa"at, and at "Ain Gev
in Israel, derive from occupational areas and furnish us with no specific
information with regard to secular or religious use. They are similar in this
respect to the bull cauldrons from the Gordion citadel.99
Distribution
Any study of movement process initially confronts the question of the

cultural-geographical origin of the material under review. Having already
briefly discussed this issue, it need only be noted that the route to the west
will have been by sea; and with a North Syrian-south central Anatolian origin
in mind, the cauldrons will have come by ship via the Mediterranean. The
ports of embarkation were presumably on the Levantine coast, but whether
at al Mina, Sukas, or other ports, remains unknown (pace all the specific ref-
erences to al Mina; v. Hanfmann 1957, 247). An overland route via Phrygia to
the East Greek coast (the normal route for Phrygian material) and thence
by sea to the Greek sites, cannot be excluded (e.g., Winter 1988, 211), but
this view implies that the North Syrians either needed an intermediary or
desired to use the long overland route when the Levantine ports were close
and accessible.
Was the movement of the cauldrons from the workshops to the ports of
embarkation part of a targeted political-social process or was it an entre-
preneurial activity? Are any patterns discernible in the movement? I think it
is possible to make the inference that the cauldrons with attachments were
considered to be valuable objects, not bric-a-brac available for causal con-
sumption, as plain cauldrons might have been. They evoke not merely a sub-
jective aesthetic quality, but a sense of spiritual conception and planning,
as well as an appreciation of the recruitment of skilled artisans who ham-
mered the cauldrons and sculpted and cast the attachmentsa complex,
labor-intensive enterprise. Moreover, the addition of attachments surely sig-
nified that the cauldrons were special objects, perhaps associated with cultic
use.
This inference, and the evidence of the find spots, leads to another in-
ference: the cauldrons were symbols of power, status-enhancing objects
99 Plain cauldrons are depicted being used in an apparently secular banquet scene in the
time of Sargon II (Herrmann 1966a, 3 f., n. 7, fig. 5), but it does not necessarily follow that
cauldrons with attachmentswhich may have given a charged value to the cauldronswere
likewise used in the Near East or in Greece.
made for and acquired both by people of rank and the gods.100 Because of the
lack of data it is not known whether a different social, religious, or political
status was accorded respectively to siren and bull attachments and if so,
whether this affected the cauldrons destinations (or Greek perceptions).
The social and economic organization of the North Syrian kingdoms
and their neighbors and royal relationships to workshops remain unknown,
although we may assume the workshops were not totally independent of
state control. But we do know that royal giftgiving had an ancient history
in the Near East, including North Syria (Winter 1988), and royal gift giving
signifies some state control of artisans and workshops (Winter 1988, 205;
Zaccagnini 1983).
North Syrian cauldronstwo with four sirens from Tumulus MM (query:
were four sirens of higher social and religious value than two sirens?) and
possibly the two with bull attachments from Tumulus Warrived at Gor-
dion in the 8th century bc. While K. Sams (1979, 46) accepts the possibility
that a number of other North Syrian objects that reached Gordion at this
time were the result of a kind of diplomatic protocol between Phrygia and
North Syrian heads of state, he asserts that the siren cauldrons there were
trade items. I would posit, however, that given both their deposition in a
royal tomb and their high status value, it is tenable to conclude that the
movement of the cauldrons to Phrygia resulted from diplomacy, not trade.
They surely represented a gift, part of an exchange, from a North Syrian king,
and possibly are material reflections of an alliance, like that between Phry-
gia and Carchemish in 717bc, or of another (earlier?) unrecorded alliance (v.
Sams 1979, 46; Winter 1988, 209).
Relevant, albeit circumstantial, evidence supports this interpretation.
Tumulus MM also contained two extraordinary animal-headed situlae that
probably arrived there from either Assyria or North Syria (v. Muscarella,
forthcoming). Textual evidence exists in the Near East that animal-headed
vessels were a standard diplomatic gift in the 2nd millennium bc (Dunham
1989).101 While we lack first millennium textual data on this custom, it is not
improbable that it continued, and that the Gordion situlae were royal gifts to
the Phrygian king. On Assyrian reliefs of Sargon II situlae of exactly the same
form are depicted being dipped into cauldrons in order to obtain liquid for
100 See Coldstream 1983b, 201, who argues that Greek terracotta vessels found on Crete
and Cyprus were of high prestige value and were gifts from Greek noblemen (208). Kolaios
dedication of a griffin cauldron to Hera is a good example of this feature.
101 Unfortunately, I did not know of these texts when I wrote Muscarella forthcoming
although I did mention the issue of gift-exchange.

pouring into cups. This is precisely the combination of vessels recovered in

Tumulus MM, but here bowls, a Phrygian preference, are the recipients of
the liquid rather than cups, and the cauldrons have attachments. May not
the cauldrons and situlae constitute an ensemble that arrived as a royal gift
to be used for a royal banquet or for a ritual ceremony?102
Did diplomatic gift-giving play a role in the diplomatic-social relations
between Greece and the Near East? As I have discussed elsewhere (1989,
333ff.), it was not for nothing that Midas dedicated his own throne to the
god at Delphi.103 The object dedicated was the supreme symbol of power,
a status-enhancing giftfor the donor as well as for the receiver. Ancient
texts reveal that royal gift-giving involved reciprocity, that in the deed ex-
change was inherent. The gift of a throne suggests the expectation of sig-
nificant return: most probably a propitious oracle and the good will of a
Greek deity, and surely at the same time Greek recognition of Midas power
and wealth. Similar reasons prompted Croesus gifts in the 6th century bc to
Delphi and the Amphiareion (Herod. I. 4655). And one may posit that pro-
tocol and caution required Midas to send the throne with Phrygian envoys,
no doubt with the cooperation of friendly East Greek shipsjust as later
Croesus sent Lydians to Greece, rather than entrusting such a gift to Greek
intermediaries, especially since the oracles answer was eagerly anticipated.
Correct understanding of the Midas dedication as a gift exchange gen-
erates a model that might explain how other material as well may have
reached Greece.104
Two separate events are before us, one source-documented (the Midas
gift), the other an inference from excavated material (the Gordion caul-
drons). Both involve Phrygia and North Syria and both presumably are to
102 Simpson 1986, 47. This view would support a North Syrian origin for the situlae (Mus-
carella forthcoming). It is also possible that the cauldrons and situlae arrived at Gordion
separately, as gifts from different kings, and were there used together.
103 The throne was most probably made of inlaid wood (Muscarella 1989, 334). Would
Midas have sent other wooden furniture to Delphi, like that from Tumulus MM, wood which
would no longer be preserved? To date no Phrygian material has been excavated at Delphi:
without the historical record we would not know of the Phrygian dedication. Could this gift
have played a role in the genesis of the rich-Midas legend?
104 For Phrygian material recovered in Greece see Muscarella 1967b, 59ff., and a revised
survey in 1989, 337 ff. In this latter work (339) I reported that two unpublished XII, 13 fibulae
with double pins were seen by me in the Olympia museum. They had in fact already been
published by Philipp 1981, 310f., pl. 69, nos. 1115, 1116. Philipp is correct that in Muscarella 1967b,
32, n. 35, I was referring to these two examples; when no. 1115 was first published in 1890 (as
no. 370) the double pin was not shown, and thus it was cited by me, ibid. 31, n. 31a, as a single-
pin example.
be seen in the context of royal gift-giving and thus are formally interrelated.
Phrygia and North Syria maintained political and cultural contact (Sams
1974, 181ff.; idem 1979, 45f.). Now, while the Phrygians were knowledgeable
about Greek affairs (Muscarella 1989), we do not know what the North
Syrians knew of the Greek sphere. Here archaeologists lack information
on crucial questions: did the manufacturers or the commissioners of the
cauldrons know their destination (Greek sanctuaries), and did the Greeks
know their source? But if gift-giving was the mechanism in force, then we
would infer that the Greeks did know the source(s), and that some Syrians
were informed about Greece.
Thus to invoke the Phrygian model to explain the presence of oriental
cauldrons in Greece is to assume that the North Syrians knew something
of Greeks and their sanctuaries, either indirectly through contacts with
the Phrygian court, and/or through contacts, commercial or diplomatic,
with Greeks on the Mediterranean coast. Furthermore, it assumes that they
desired to communicate with Greek deities.
Although an assumption and not an inference, this conception is by no
means untenable, especially given the quantity of North Syrian material in
Greece (Muscarella 1970, 116ff.; Winter 1988, 196 ff., 207); and it further allows
us to perceive a connection between the two events mentioned above. If one
or more North Syrian kings gave cauldrons as gifts to the Phrygian court,
then he/they might also give gifts elsewhere; and if one oriental king gave a
gift to a Greek deity, then other kings may well have done the same, for the
same general reasons.
Inferences and assumptions do not necessarily add up to historical real-
ity, but I argue that the Phrygian model with its foundation in Herodotus
could legitimately be adduced to explain how perhaps some oriental caul-
drons reached Greece: as gift-dedications.105 It would be rash to go beyond
the some, because obviously other mechanisms may equally have been
involved.
105 Herodotus (I. 14) states that after Midas dedication at Delphi a gap of some 3060 years
ensued before another oriental king, Gyges (died ca. 650 bc), dedicated an object. This claim
is not a problemfor Herodotus is of course speaking about Delphi, not other sanctuaries;
and even for that site he may have been reporting only about important kings remembered
hundreds of years after the fact by his local informers. Note that Herodotus account informs
us that at least in some cases the Greeks were aware of the specific oriental sources of
the imports in the 8th and 7th centuries: see also Herod. I. 4555 for 6th century Lydian
imports.
I. Winter (1988, 210f., 212) has argued that al Mina was an outlet for goods
from North Syria, and that the presence of Greeks there indicates that North
Syrian goods were being distributed as the products of trade, as well as the
result of booty, tribute and giftgiving. This view of the apparent mecha-
nisms allows for gift-giving (but with no specific examples indicated). With
her broad claims I have no quarrel in general, but they beg a number of
questions, questions that I doubt can be answered meaningfully at present:
which goods that reached the Greek sphere were gift, which trade items?
Were some cauldrons with attachments gifted, others traded, and if so, was
trade part of their manufacturing strategy? Who within North Syria (and
elsewhere) sold and profited from the sale, the court or private merchants, or
both? There are other questions: where were these cauldrons sold, in North
Syria, on the coast, or in Greece? To whom were they sold, to other Near East-
erners and/or to Greeks? We may also ask why are the cauldrons not found
in private Greek tombs? And what did the Greeks, if they purchased caul-
drons, give as payment?106 Not knowing the answers to all these questions
precludes definitive statements about the movement process.
Moreover, to assert that either Greek or Phoenician ships transported
the cauldrons west tells us nothing about how and why they got to these
ships, whether for trade or giftgiving, nor who owned them at the time of
embarkation. This is why it is insufficient to suggest that Greeks brought the
cauldrons west (Herrmann 1966a, 149; Cold stream 1977, 360; Winter 1988,
211) or that Near Easterners carried them (Dunbabin 1957, 35 ff., 49; Muhly
1970, 48f.) without confronting the questions raised here. Winter (1988, 212)
aptly cites a reference to a 7th century bc merchant from Carchemish, but
the reference does not inform us of his responsibilities or of his wares.
Trade, then, while it may be posited as a possible process of distribution,
is not readily demonstrated, even if we follow the discussions in scores of
books and articles on trade patterns and models. Thus, while I have found it
possible to ascertain a Phrygian-Midas model for the movement of certain
high-class objects, I have not been able to comprehend a trade model
which failure, however, does not exclude trade from being considered as a
possible process.107
106 Coldstream 1982, 265 suggests that Greek silver was used as payment.
107 That a successful Greek merchant, Kolaios, could dedicate a cauldron with griffin
protomes at a Greek sanctuary (Hera on Samos) in the 7th century bc is indicated by
Herodotus (IV.152). But this act conforms to an old Greek custom: a Greek commissions and
then dedicates a Greek-made cauldron to a Greek deity. Thus I believe that this knowledge
cannot reasonably allow archaeologists to invoke a Kolaios modelwhich would obtain
To my knowledge only Filippakis et al. (1983) have suggested that ori-

ental immigrant artisans, and subsequently their Greek apprentices, man-
ufactured the cauldrons in Greece. It was earlier proposed by Dunbabin
(1957, 41, 49) and later by Coldstream (1977, 56, 70, 100, 103, 287 ff., 358; 1982,
266ff., 271) that Phoenicians and other orientals worked at Athens and on
Crete in the 9th and 8th centuries bc, some fleeing the conquest of Sar-
gon II. Muhly (1970, 48f.), discussing the same Cretan objects cited by Cold-
stream as locally made, sees them to be imports brought by Phoenicians,
although he too would accept the presence of Phoenicians on the island
(42). Boardman (1970, 18ff.; 1980, 57ff.) also believes that Orientals, North
Syrians, arrived in Crete and there produced artifacts. For his conclusion he
cites archaeological evidence in the form of 7th century burial customs that
seem to be the same at Afrati and at Yunus, near Carchemish.108 And more
for Greek dedications aloneto justify the conclusion that oriental cauldrons were also
commissioned/purchased by Greeks in the 8th or 7th centuries bc for dedication at Greek
sanctuaries.
Nor ought we be quick to accept the Denkmodell of H. Kyrieleis in Kyrieleis and Rllig
1988, 3761, with a philological discussion by W. Rollig (6275). Here Kyrieleis publishes a
superbly decorated North Syrian bronze horse plaque found in a 6th century context on
Samos in 1984. It is inscribed in Aramaic with the name of Haza"el and is thus dated to
the late 9th century bc (49; Rllig, 62 ff., 71 ff.). Reference is also made to one of two bronze
North Syrian blinkers excavated around 1900 at Eretria on Euboea (Muscarella 1970, 116,
fig. 10) that after cleaning revealed the same inscription as on the Samos piece discussed
(49, 55 f.; Rllig, 69 ff.). A. Charbonnet earlier (1986, 117, 140ff., fig. 3) published a different
interpretation of the Eretrian inscription and accorded it an 8th century date (see below).
Kyrieleis presents two crucial assertions: the two objects arrived on Greek soil in the 9th
century bc as a result of trade, or better, as gift-exchange transactions between enterprising
Greek merchants and a local ruler; and the objects were later dedicated at the sanctuaries
by Greeks, not by Near Easterners (56 ff., 59, 61). Furthermore, as a consequence of these
assertions he postulates what are surely major historical conclusions concerning early Greek-
Near Eastern contacts and the date of the creation of the Greek alphabet (58ff.). But we are
dealing here with intuitive guesses; we are not able to accept it as archaeological explanation
for events that no archaeologist is capable of documenting. We do not know when the objects
reached Eretria and Samos (8th century bc?), or the political and cultural background of the
donors. And I strongly argue that the two objects under review may not, cannot, be presented
legitimately as evidence for early, 9th century bc, oriental writing on Greek soil. Note that
Charbonnet 1986, 117, 143, 145, 157, believes that the Eretrian plaque arrived at the site in
the 8th century bc and that it was dedicated by an Aramaean who inscribed the plaque on
the spot: also guesswork, in part based on a mis-translation of the text. It is also relevant to
recall here that in his discussion of the Hundehalter figurines that were excavated on Samos,
Kyrieleis 1979, 44 f., 46, 48, argued that surely Near Easterners, not Greeks, dedicated these
oriental objects to Hera. Note that both Strm and Rllig, infra, 48, 97, accept an 8th or 7th
century date for the deposition of the two pieces.
108 But the alleged foundation deposit cited from a tomb at Khaniale Tekke (Board-
man 1970, 15; idem 1980, 57; Coldstream 1977, 100; idem 1982, 267) does not clearly reflect a
recently J. Shaw (1989) has published strong evidence for the presence of
Phoenicians on Crete, an eastern sanctuary at Kommos.
None of these scholars posits immigrant craftsmen to explain the pres-
ence of oriental cauldrons in Greece, and, in fact, Coldstream and Boardman
accept the cauldrons as imports. One might be tempted to suggest that those
bull attachments that scholars find difficult to call either Greek or orien-
tal are hints of Orientals working in Greecereflecting their adjustment. But
this view ignores the reality that the problems here may be an artifact of
modern perceptions and analyses. There is ample evidence that in the Near
East throughout its history artisans moved either voluntarily or under orders
(Zaccagnini 1983), but it would be infelicitous to suggest that oriental arti-
sans came to Greece to manufacture cauldrons for dedication.
Near Eastern custom. Frankenstein 1979, 274 ff., argues that the mechanism of movement of
Phoenician goods west was trade, not immigrant craftsmen.
Section Two
Artifacts
chapter twenty-five
FIBULAE REPRESENTED ON SCULPTURE*
Archeologists have long recognized the value of fibulae as a chronological

guide and as an indication of ethnic or cultural movements.1 This brief study
will present ideas about several other ways in which the scholar may gain
significant information from a study of these small objects. Its basis will be
those fibulae carved on stone in the form of reliefs, friezes, and plaques.
The best known example of a fibula represented in art east of Syria is the
one portrayed on the Khorsabad frieze of the last years of the eighth cen-
tury bc2 (Pl. II, Fig. 1). It is a Phrygian fibula although neither the typical
horned catch,3 nor the pin, which is generally not shown on sculpture, is
depicted. The lack of a horned catch is a fault or oversight of the sculptor
who, in my opinion, intended to show a Phrygian fibula of the type cat-
egorized by Blinkenberg as XII,7.4 This type is very common at Phrygian
centers, particularly at Gordion where about seventy-nine examples have
been found to date in tombs of the late eighth century bc. It is without doubt
a Phrygian product.5
The fibula on the frieze is worn by a man who is part of a procession of
tribute bearers, all of whom wear the same type of clothing, indicating, it
* This article originally appeared as Fibulae Represented on Sculpture, Journal of Near
Eastern Studies 26, no. 2 (1967): 8286.

1 G. Kossina, Die deutsche Vorqeschichte, Mannus Bibl., IX (1921), 48; K.R. Maxwell-Hyslop,
Notes on Some Distinctive Types of Bronzes from Populonia, Etruria, Proc. Prehist. Soc., 1956,
p. 131.
2 P.E. Botta, E. Flandin, Monument de Ninive, II (Paris, 1849), Pls. 103, 106 bis. Some
females represented on objects from Luristan wear what appears to be long straight pins
at the shoulders, E. Porada, The Art of Ancient Iran (New York, 1965), p. 89, and Fig. 60.
These representations are very similar to the straight pins worn by some Greek ladies on
the Franois vase, H.L. Lorimer, Homer and the Monuments (London, 1950), p. 379, Fig. 54.
Note that it is possible the goddess with mirror represented on the Hasanlu gold bowl wears
straight pins on her shoulders, Porada, op. cit., p. 101.
3 The horned catch is one of the most characteristic features of Phrygian fibulae, see
Oscar White Muscarella, Phrygian Fibulae from Gordion, 1965, an unpublished dissertation
in the University of Pennsylvania Library (hereafter Phrygian Fibulae).
4 Fibules Grecques et Orientales (Copenhagen, 1926), pp. 213f. (hereafter Fibules); Phrygian
Fibulae, pp. 29 ff.

5 The fibulae from Gordian will be published by the author in the near future.
768 chapter twenty-five
would seem, that they all belong to one national group. Yet the garment
of only one of the bearers is shown open its whole length, in front view,
the two borders clasped by a fibula. In order to show the manner in which
the fibula is worn, the sculptor has depicted this man with arms raised
supporting a sack on his shoulders; the other gift bearers have their arms
before them. Thus it would appear that there has been a deliberate attempt
to call attention to the fibula, and I suggest that the sculptor wished to
indicate that Phrygians were passing by.6 We gain here an insight into the
practice of an ancient sculptor, or his supervisor, who wished to identify a
specific group of people, the Phrygians, possibly for a non-reading public. He
simply chose to illustrate one of their most characteristic products, a fibula.
If this interpretation is correct, I suggest that the procession represents
gift bearers from King Midas in the year 710/709bc, the year Phrygia was
attacked from Cilicia by Sargons agents.7 The frieze incidentally proves that
the Type XII,7 fibula was used in the late eighth century bc.
An exact parallel for the use of a fibula as an indicator of nationality is
provided at Persepolis by the Apadana reliefs and by the dais of Artaxerxes I
in the Throne Hall. On the Apadana walls many groups of allies and sub-
jects of the Persian king are represented bringing gifts, perhaps for a new
years festival. Only one of these groups is depicted wearing fibulae (Pl. III,
Figs. 2, 3), and only two of the fourteen men supporting Artaxerxes dais wear
fibulae.8 There can be little doubt that the sculptors were aware that certain
peoples characteristically wore fibulae and so represented them. Therefore,
we may assume that the ancient passerby, unlike the modern scholar, eas-
ily recognized the nationality of the fibula wearers. Barnett claimed that the
fibula wearers were Phrygians; and Bittel supporting E. Schmidt, maintained
that they were from Central Anatolia, in particular from Alisar, even though
he could find no clearly similar examples there. One type of fibula repre-
sented on the Apadana reliefs, has a ribbed arc, a spring and a hand catch
(Pl. IV, Fig. 3); it is apparently of the same type as, or similar to, those on
6 R.D. Barnett, Early Greek and Oriental Ivories, JHS, LXVIII (1948), 9.
7 D.D. Luckenbill, Ancient Records of Assyria and Babylonia, II (Chicago, 1927), 41, 43.
8 E. Schmidt, Persepolis, II (Chicago, 1953), Pls. 35 A, B, 108, 109, 112. See also R.D. Barnett,
Persepolis, Iraq, XIX, 1 (1957), 67; K. Bittel, Fibeln in Perspolis, Vorderasiatische Archologie,
Ed. by K. Bittel (Berlin, 1964), pp. 39 ff., Pl. 9. It is interesting to note that the arcs of the
Persepolis fibulae are worn both upright and pendent. This demonstrates that there was a
choice. Some fibulae would look better with the arc upright, viz. Iraq, XXII (1960). Pl. XXX,
p. 3, a gold fibula from Ziwiyeh; and most of the fibulae from Boeotia, in Greece, Fibules,
pp. 187 ff.
fibulae represented on sculpture 769
earlier North Syrian reliefs (infra). Another type of fibula, represented both
on the dais and on the Apadana reliefs, has a plain arc and what seems to
be a hinge rather than a spring (Pl. III, Fig. 2). Fibulae of both types, those
with hand catch (Near Eastern), and those with hinges, were in common use
in Western Iran and in the South Caucasus.9 The hand catch and hinge are
crucial in this discussion, for neither was employed extensively in Anatolia
although some imported pieces have been found.10 Therefore I suggest that
the fibulae wearers should tentatively be identified as an Iranian or perhaps
a South Caucasian people. The dress of the group does not contradict this
opinion, for they are dressed like the other groups on reliefs at Persepolis,
one of which Schmidt has identified as Medes.
We have presented two examples of fibulae employed by sculptors to des-
ignate particular national groups, that at Khorsabad and those at Persepolis.
The well-known and often illustrated relief at Ivriz presents us with a differ-
ent situation in which the fibula is not used as an indication of nationality.
On this relief, King Urpallu of Tabal (fl. 738 bc) stands before a god.11 Urpallu
wears elaborately decorated clothing. The outer garment is secured by a
Phrygian fibula of a variety similar to examples found at Gordion and Ankara
in Phrygian contexts.12 This particular variety has different arc forms and
is characterized by a double pin masked by a rectangular transverse plate,
rather than the usual single exposed pin. The plate is detachable and serves
to shield the pins; two tubes sheath the points (Pl. IV, Fig. 4). Fibulae with
9 Pace Bittel, op. cit., pp. 40, 41, n. 16. See n. 10. Also in 1964 at Hasanlu a Pd. II (fifth-fourth
century bc) cist tomb yielded a hinge type and a spring fibula found on the same skeleton
(unpublished).
10 Oscar White Muscarella, A Fibula from Hasanlu, AJA, LXIX 2 (1965), 234, nn. 11, 12;
p. 235, n. 20. Add Alishar, OIP, XIX, 266, Fig. 353; XXIX, 439, Fig. 493; XXX, 111, Fig. 106, 2041;
180, Fig. 201, d25, e776. A few imported Near Eastern fibulae have been found in the later
levels above the destroyed Phrygian city at Gordion, Phrygian Fibulae, Appendix D, pp. 220f.
In 1964 some Near Eastern and hinged types of fibular were found at Hasanlu in late seventh
century and fifth and fourth century contexts (Periods III and II), Oscar White Muscarella,
Hasanlu 1964, BMMA, Nov. 1966, p. 135, Figs. 37, 38. It should be noted here that the fibula
on the Khorsabad relief discussed in the text above has nothing to do with the fibulae on the
Persepolis reliefs. As I hope the texts makes clear, one is Phrygian, the others Near Eastern,
see nn. 3 and 4 supra. Recently W. Culican in The Medes and Persians (New York, 1965), p. 99,
erroneously calls the Persepolis fibulae Phrygian, perhaps following Barnett Iraq, 1957, p. 67.
11 The most recent and best illustration of the relief is to be seen in E. Akurgal, The Art of
the Hittites (New York, 1962), Pl. XXIV, Fig. 140.

12 H. Zubeyr, Ankara Gazi Orman Fidanliginda Bulunan Eserler, TTAED, I (1933), 15, nn. 9,
10. Only a few of the Gordion examples have been published, viz. AJA LXII, 2 (1958), Pl. 25,
Fig. 20; Expedition, VI, 2 (1964), lower left figure on p. 37. See also Phrygian Fibulae, p. 7, Fig. 1;
22; 33, Figs. 8, 9; 63, Fig. 16; 75; 127.
a fixed transverse plate, which has undoubtedly been copied and modified
from the Phrygian examples, have also been found.13 The fibula represented
on the Ivriz relief has an arc of XII,9 type, according to Blinkenbergs classi-
fication. Although no examples of this particular type have yet been found
with double pins and detachable transverse plates, many with single pins
do occur in Phrygian contexts of the late eighth and seventh centuries bc at
Gordion and elsewhere14 (Pl. IV, Fig. 5).
Since Urpallu was a king of Tabal, an eastern neighbor of Phrygia, we
cannot call Urpallu a Phrygian simply because he wears a Phrygian fibula.15
We may assume in this case that the fibula was an import to Tabal worn by
the king as a piece of jewelry and so depicted on the relief by the sculptor.
Another representation of King Urpallu exists on a relief found in frag-
ments at Bor, northeast of Ivriz, and now in the Istanbul Museum (No. 837).
As exhibited at present, the relief has been restored and a Phrygian fibula is
worn over the chest. The reason for adding a fibula to the restored area was
recently explained to me in a letter from Bay Necati Dolunay, Director of the
Archeological Museums of Istanbul: the authorities decided to use the Ivriz
relief as a model for restoration and, since Urpallu wears a fibula there, it
was decided to copy it on the Bor relief. Whether the latter originally had
a fibula is not known. Through the kindness of Bay Dolunay, I illustrate the
relief both before and after restoration (Pl. V, Fig. 6).
Several examples of fibulae on reliefs have also been found in North
Syrian cities, not far from Ivriz: one fibula from Zincirli on a relief of a queen,
one on a female who is part of a royal couple from Marash, and two others on
a stone fragment of clothing from Carchemish (Pls. VIVIII, Figs. 79).16 All
these fibulae are of the same type, semicircular with a ribbed arc; they are
found in many parts of the Near East including Iran, from the eighth century
and later.17 They were commonly worn in daily life. When represented on the
13 Fibules, pp. 217 f., Type XII, 10; U. Jantzen, Phrygische Fibeln, Festschrift fr Friedrich
Matz (Berlin, 1962), pp. 39 ff., Pl. 8, Nos. 1, 2, Pl. 11, Nos. 13; D.G. Hogarth, et al., Excavations
at Ephesus (London, 1908), pp. 151 f., called a handle in the publication; R.M. Dawkins, et al.,
The Sanctuary of Artemis Orthia at Sparta (London, 1929), Pl. 83 c; Phrygian Fibulae, pp. 48ff.;
Alisar, OIP, VII, 96, Fig. 76, 2855.
14 See n. 5. Also J.M. Birmingham, The Overland Route Across Anatolia, Anatolian Studies
XI (1961), 186 f.
15 Cf. Barnett, op. cit., pp. 8 f.; Phrygian Fibulae, pp. 145ff., nn. 8 and 9.
16 F. von Luschan, Ausgrabungen in Sendscherli, IV (Berlin, 1911), Fg. 236, Pl. 54; E. Akurgal,
Spthethitische Bildkunst (Ankara, 1949), Pl. XL; C.L. Woolley, and R.D. Barnett, Carchemish,
III (London, 1952), 240, n. 2, Pl. B64C.
17 As D. Stronach already pointed out, The Development of the Fibula in the Near East,
Iraq, XXI, 2 (1959), 189; we add Marash to his list. The relief fibulae under discussion have
reliefs under discussion, they have no significance beyond the fact that the
sculptor was representing accurately the daily dress of his subjects. Their
value for the archeologist is chronological for they are well dated artifacts of
the late eighth and seventh centuries bc.
Fibulae are often represented on other groups of sculptures from the
Near East, and in these cases with interesting implications. These sculp-
tures are exorcizing plaques or amulets used in rituals to ward off sickness.
The amulets are decorated with scenes in relief; one of these scenes has a
demon as a central figure, and various objects, including a fibula, are often
represented in the neighboring field (Pl. VIII, Fig. 10).18 These objects were
offerings, votive gifts, for the demon represented on the amulet.
Classical archeologists are familiar with the fact that fibulae played a
role as votive gifts, as offerings to a deity, in sanctuaries of Greece and Asia
Minor.19 Exactly what this role was still remains obscure. Little light is shed
on the subject by Herodotus etiological story in Book V, 87, 88. He describes
how the Athenian women murdered the sole survivor of the war with Aegina
by means of their peronai, and subsequently dedicated their peronai in
sanctuaries at Argos and on Aegina. It is not clear in these passages whether
Herodotus is writing about straight pins or fibulae. However, since he refers
to the objects being dedicated at temples, and since almost every Greek
sanctuary of the first millennium bc has yielded large numbers of fibulae,
I suggest that he is referring to fibulae rather than to straight pins. It is
also possible that Herodotus has joined two seperate stories, one concerned
with the murder, where straight pins may have been the weapon, the other
concerned with votive offerings, where fibulae were used.
incorrectly been called Asia Minor (i.e. Phrygian) types: Fibules, p. 29; Lorimer, op. cit., p. 354;
E. Akurgal, Chronologie der Phrygischen Kunst, Anatolia, IV (1959), 116.
18 J.B. Pritchard, The Ancient Near East in Pictures (Princeton, 1954), Figs. 658, 660; H.H. von
der Osten, A Babylonian Amulet, BMMA, XIX (1924), 145 ff.; Zwei Neue Labartu-Amulette,
AfO, IV (1927), 89 ff.; F. Thureau-Dangin, Ritual et Amulettes contre Labartu, RA, XVIII
(1921), 172 ff., Pl. 1. For a more extensive bibliography on these amulets see H. Klengel, Neue
Lamastu- Amulette aus dem Vorderasiatischen Museum zu Berlin u. dem Br. Museum, MIO,
VII (1959), 334 ff.; for fibulae, pp. 338 f., Fig. 1a, b; pp. 344 ff., Fig. 6a, b, and n. 29. See also by
the same author Weitere Amulette gegen Lamastu, MIO, VIII (1960), 24ff., for fibulae p. 27,
Fig. a, b. Klengel alone seems to be aware of the fibulae representations but comes to no
conclusions about them.
19 Fibules, pp. 19 ff., 35, 106, 182, 193 f.; R. Hampe, Frhe Griechische Sagenbilder (Athens,
1935), pp. 3. 90 ff.; Lorimer, op. cit., pp. 249 f.; A.M. Tallgren, Caucasian Monuments, ESA, V
(1930), 178 f. A fine archeological example of fibulae used as votive gifts may be seen in the
gold and silver sheetmetal types found in the basis at Ephesus, Hogarth, op. cit., p. 98. Pl. V, 6,
117, Pl. XI, 12. These objects could not actually be used to fasten clothing and must have been
dedicated as fibula simulacra.
The fibulae found in graves in Greece and surrounding areas in my opin-

ion may also have been votive offerings (Charons penny!). It is known from
representations on reliefs, terracottas, and vase paintings20 that no great
number of fibulae were worn in daily life. How then can we account for
the fact that in many graves from different areas, larger numbers of fibulae
have been found?21 One obvious explanation is that extra fibulae, along with
other goods, were placed in the grave for the sole convenience of the dead,
that is for personal use in the netherworld. Yet, since fibulae were employed
as votive offerings in a temple sanctuary, they might then also have been so
used in a tomb. In this case they would have been intended as gifts to be
presented by the individual to the deities he might expect to meet.
When we return to the Near East, we find in some Sumerian texts recently
discussed and summarized by S.N. Kramer22 some relevant information, the
archeological implications of which are quite significant for our theory
and also for the interpretation of tomb material in general. The texts are
concerned with dead kings who upon entry into the netherworld present
various deities with giftsanimals, weapons, ornaments, etc.precisely
those very objects placed in their graves. Although the texts are early, the
concepts expressed are by no means to be regarded as inapplicable to later
periods and different areas. It should be noted, however, that in classical
texts concerned with funereal offerings there is no suggestion that gifts
offered at a tomb were for a deity; rather they were for the honor of, or use by,
the dead individual.23 I suggest that the question remain open and that the
possible votive use of fibulae placed in tombs should be further investigated.
One should be concerned, for example, with the geographical extent of such
practices.
20 Besides the examples discussed in this study see Fibules, pp. 28ff. Some more recent
finds include G. Buchner, Figurliche bemalte Sptgeometrische Vasen aus Pithekusai u.

Kyme, RM, 19531954, 19601961, 52, Fig. 3; J. Boardman, The Cretan Collection in Oxford
(Oxford, 1961), p. 104, No. 479, Pl. XXXVI; J. Charbonneaux, Les Bronzes Grecs (Paris, 1958),
pp. 70 f., Pl. XVI. This is not a definitive list.
21 For example, at Gordion the Phrygians placed many fibulae in tombs of the eighth and
seventh centuries bc, C. and A. Koerte, Gordion, Ergebnisse der Ausgrabungen in Jahre 1900,
JdI, Supplement No. 5 (Berlin, 1904), pp. 45 ff., 76 ff., 101 ff.; AJA, LXI, 4 (1957), 327; AJA, LXVII, 2
(1958), 149; AJA, LXIV, 3 (1960), 227 ff.; Hampe, op. cit., p. 4; Notz. degli Scavi, 1880, p. 77; 1895,
p. 167; E. Gjerstad, Early Rome, II (Lund, 1956), 89. There is no strong evidence forthcoming
from Near Eastern graves to demonstrate that fibulae were placed there in large numbers.
22 Death and Netherworld According to the Sumerian Literary Texts, Iraq, XXII (1960),
59 ff. These texts deserve special study by anthropologists and archeologists who are inter-
osted in why objects were placed in graves.
23 Thucy. iii, 584; Herodotus v, 92; Euripides Orestes, 115ff., 1453ff.; Sophocles Elektra, 431ff.;
Plutarch Aristides. xxi; Tacitus Annals iii. 2; Vergil Aeneid xi. 195f.
The exorcising amulets, which were the starting point of this discussion
on fibulae as votive objects, document the fact that fibulae did have religious
value connected with at least one deity in the Near East during the seventh
century bc. And quite relevant in this context are some interesting fibulae
found in Palestine and Iran, probably of the late eighth and seventh cen-
turies bc.24 These fibulae have a demons head (a Pazuzu?) at one end of the
arc and a female (a goddess?) or bird at the other end. Note that the relation-
ship of the demon with a fibula also occurs on the amulets. Is it not possible
that these fibulae have a votive or at least an apotropaio, function? I suggest
therefore that we may conclude on the basis of the amulets and these fibulae
that in the Near East as well as in the Greek world fibulae had some special
religious function. Whether the very same value was applied in both areas
cannot be known at present (especially with respect to fibulae in graves),
and more research will have to be done before one may state in which area
the idea originated. Nevertheless, it is suggested that the evidence presented
here has brought forth information about a hitherto unrecognized relation-
ship between the Greek world and the Near East in the early first millennium
bc.
24 R. Ghirshman, La Fibula en Iran, Iranica Antiqua, IV, 2 (1964), Pl. XXV, 1315, also
Pl. XXIV, 11; R.S. Lamon, and G.M. Shipton, Megiddo, I (Chicago, 1939), Pl. 71, No. 72. I owe
the Megiddo reference to Professor Edith Porada. These Iranian fibulae were not known to
me when I wrote A Fibula from Iran; see n. 10.
Fig. 1. Frieze from Khorzabad. Late Eighth Century bc.

Fig. 2. Persepolis, Apadana, Northern Stairway, Early Fifth Century bc.
Fig. 3. Persepolis, Apadana, Eastern Stairway. Early Fifth Century bc.

Fig. 4. Fibula from Tumulus MM at Gordion. Ca. 725700bc.

Fig. 5. Fibula from Tumulus MM at Gordion. Ca. 725700 bc.

Fig. 6. Relief from Bor, Before and After Restoration. Istanbul No. 837.
Fig. 7. Relief from Zincirli. Eighth Century bc.

Fig. 8. Relief from Marash. Eighth Century bc.

Fig. 9. Fragment of Relief at Carchemish. EighthSeventh

Century bc. Courtesy of the Trustees of the British Museum.
Fig. 10. Limestone Amulet from Nimrud. Seventh Century bc.

Courtesy of the British School of Archeology in Iraq.
chapter twenty-six
PHRYGIAN OR LYDIAN?*
The purpose of this paper is primarily to publish a remarkable gold fibula

allegedly found in Anatolia. At the same time, I wish to discuss briefly some
archeological problems concerning the fibula and other material found in
Anatolia between the late eighth and sixth centuries bc.
The fibula (Pls. III, Figs. 15) is presently in a private collection in the
United States.1 Because it was purchased from a dealer in antiquities, we
have no objective record of its provenience; we may, however, accept the
information that it was found in Anatolia.
The fibula is a well-known Phrygian type with characteristic moldings on
the arc and a horned catch; it may be recognized as a Type XII, 14 example
of Blinkenbergs Asia Minor group.2 The arc is a hollow tube carefully curved
* This article originally appeared as Phrygian or Lydian? Journal of Near Eastern Studies
30, no. 1 (1971): 4963.

1 The height of the assembled fibula is 5.2 cm., width 5.5 cm., length of double pins 3.3cm.,
l. of lock-plate 5.2 cm., w. 1.4 cm., diam. of hemispheres on lock-plate ca. .2cm., l. of chain 4cm.
Karat tests show a variation: the obverse of the lock-plate: 18K, one of the curved pieces in
the catch: 18K, the tube/tongue of the double-pin mechanism: 16K, one of the hollow balls
at the end of the chain: 20+ K. I wish to thank the owner of the fibula for lending it to the
Metropolitan Museum of Art and Miss Kate Lefferts and Mr. Ed Rowe of the Conservation
Department of the Metropolitan Museum who, among others, generously gave much of their
time in discussing the fibula.
Minute fragments of black silk were seen by me under a microscope: on the point of one
of the scrolls on the lock plate and on the underside of one of the two prongs of the double-
pin mechanism. Mr. William Young of the Boston Museum of Fine Arts also found such a
fragment on the catch; he identified it as raw silk. I am most grateful to him for informing me
of his discovery.
Because it is fair to discuss openly and objectively all information one has about an
unexcavated object I must state here that some people think the fibula is a forgery, at
least in part. To be sure, when I first examined it I was puzzled by certain aspects of its
construction, its differences from the bronze examples known to me, the presence of a tassel,
and the decoration on the lock-plate. I discuss these features in the text. Obviously, I do not
think the object is a forgery but, if it develops that it is modern, the problems concerned
with the purchase of works of ancient art are more complex and frightening than even
now recognized. I would hope, in any event, that the comments made in this paper about
artistic continuity and about Phrygian and Lydian relationships are valid, independent of
the publication of the gold fibula.
2 C. Blinkenberg, Fibules Grecques et Orientales (Copenhagen, 1926hereafter cited as
784 chapter twenty-six
to form a near-semicircle. Five units of moldings decorate the arc in a

symmetrical fashion; each unit was added separately and held in place by
solder. The catch is separately made but permanently socketed into one end
of the arc by means of a tongue secured by solder. A double-pin mechanism
was made as a separate unit and loosely fits into the other end of the
arc by means of a tongue made in the form of a hollow tube. The unit is
detachable from the arc and seems to have been made that way originally.
The spring is stiff and without tension; its outer border is decorated with
very fine granulations. A tassel consisting of a braided chain that ends in
four strands terminating in hollow spheres is loosely attached by a thin wire
to a hole placed in a horizontal bar directly below the spring. A separately
made detachable bar or lock-plate served as a cover for the double-pin
mechanism. It consists of a strip joined at one end to a U-shaped recessed
unit that holds two tubes serving both to sheath the double pins and to
secure the catch. Two thin prongs at the other end fit into shallow grooves
behind the decorative spool placed below the catch (Figs. 1, 4). The surface
or obverse of the lock-plate is defined by a low border and a decorated area
within that is marked off by a straight strip bent to create a rectangular area.
This area is surrounded by hollow hemispheres soldered to the strip. The
marked-off area is further decorated in filigree with thin wires laid on edge
to form a scroll-like pattern with the ends of the scrolls made into arrow-like
points.
Such is the general description of the fibula. But what is extraordinary and
fascinating is the fact that each of the three basic componentsarc, double-
pin mechanism and lock-plateis created from many separate pieces of
gold ingeniously put together. The total number of pieces is more than 236
and each piece represents an individual stage in the construction.
The arc consists of a hollow tube formed from a flat piece of metal and
soldered at the seams. The catch is constructed from nine pieces of metal:
two curved pieces that form the hooks and lateral spurs or horns, projecting
from the top, and one curved piece that forms the central spine; a broad T-
shaped strip placed in the inside curve of the hooks and soldered to the three
curved pieces; two grooved spools soldered to both ends of the horns, with
two studs (or spheres) placed in these spools; finally, a rounded pie-shaped
piece of metal shields the socket hole at the end of the arc. The two end
moldings on the arc are made apparently of four grommet-like objects plus
Fibules), pp. 204 ff., 222 ff., Fig. 255; Oscar White Muscarella, Phrygian Fibulae from Gordian
(London, 1967hereafter cited as Phrygian Fibulae), pp. 24f., Figs. 6374.
phrygian or lydian? 785
three disks; these grommets are smoothly made apparently in a form. Each
of the five edges of the moldings is masked by a wire; in three cases the wire
is vertically grooved throughout, in two cases it is horizontally grooved. A
plain wire collar around the arc exists just above the moldings where solder
might have been visible. Therefore, each end molding consists of thirteen
parts. The center molding apparently consists of two grommets and two
disks, masked by three wires. In addition, a wire collar around the arc exists
at each side, making a total of nine parts. The quarter moldings are made
from two half grommets, one wire mask, and two collars at the ends, for a
total of five parts for each quarter. All the wires and collars show seams. A
red coloration is visible at all the areas where soldering occurred. The total
number of parts in the arc and moldings consists of at least fifty-five separate
parts.
The double-pin mechanism consists of two pins each soldered to a solid
vertical rectangular bar. This bar is itself soldered to a pierced horizontal
solid bar placed perpendicular to it which in turn is soldered to a spool
decorated in two zones with grooves. The spool seems to have been made
in one piece, but two thin semi-circular strips were placed over the top and
bottom of the spool (barely visible in Figs. 3, 5). The horizontal pierced bar
is soldered to the spool only at the ends, leaving the central area clear, and
leaving two shallow grooves for passage of the prongs on the lock-plate. In
addition, a very small piece of wire was placed as a reinforcement at both
corner junctures where the bar joins the spool (one of these is now missing).
The vertical bar holding the pins is also soldered at one end to the spring.
This spring is not functional and seems to have been made by twisting a strip
of wire decorated with incisions into the desired shape. Eighty-four (one is
missing) perfectly formed very small granulations were added to the edge
of the spring. Above the spring is a gadrooned disk, or spring-plate, perhaps
made in two pieces, with an additional grooved wire masking its edge. A thin
grooved strip of wire masks the join of the spring to the spring-plate (visible
in Figs. 1, 3), but we may assume that there was also solder at the join. On
the upper part of the spring-plate a short hollow tube that functioned as a
tongue was soldered; it has a lap-join that was soldered.
A braided wire chain consisting of many interconnected loops of wire was
loosely attached to the hole in the horizontal bar connected with the spool
by means of a wire loop (Figs. 2, 5). This hole has on one side, the outer
visible one, a thin wire collar. At the upper end of the chain is a free-moving
spool or grooved collar, and at a point where the chain divides into four
units is another more elaborate collar, also free-moving. This latter collar
is made up of two pieces and has wire covering two of its edges. Finally,
Plate I
Figs. 13. Gold fibula from Anatolia; anonymous

loan to the Metropolitan Museum of Art, L.69.12.
four hollow spheres cover the ends of the four braided wires; although
their collars appear to have been added, I could see no seams. Thus, we
have a total of at least thirty separate parts plus the many chain loops,
plus eighty-four granulations used in the construction of the double-pin
mechanism.
The last unit to be described, the lock-plate, has a total of sixty-seven parts
in its construction. The plate itself is formed from a strip bent up on three
Plate II
Figs. 4 and 5. Gold fibula from Anatolia; anonymous

loan to the Metropolitan Museum of Art, L.69.12.
sides and bent further at the right partially to inclose a fourth side. One
end of the plate is forked to fit behind the spool soldered to the spring; the
thin tines each had a strip of wire soldered to the rear for reinforcement
(one is now missing) (just visible in Fig. 2). A U-shaped unit, higher at one
end, was soldered to this plate and it alone consists of six pieces (Figs. 3,
4): the U-shape itself, two wires forming an arched recess that decorates
the outer face of the U-shaped recess, two hollow hemispheres each framed
with a collar placed in this outer arched recess, and seven thin strips added
at several places as reinforcements. And finally, there are the two tubes,
rolled and soldered, within the recess to hold the double pins. Although the
tubes are eight mm. in length, the pin ends penetrate only two mm. The
obverse part of the lock-plate is decorated on the outer edge with twenty-
eight hollow hemispheres that slightly vary in size and wall thickness; on
several of them may be seen a minute hole visible through a microscope.
There is also a zone, formed by one piece of wire bent to make a rectangle,
that incloses the filigree scroll-like motif composed of nineteen pieces of

wire. These wires were hand-cut and are rounded at the top, flat on the
bottom.
This prolonged description is given not to bore the reader nor to clutter up
his mind with painful details, but rather it is given to illustrate the intricate
and elaborate details involved in the workmanship. It is also hoped that such
a detailed description will help scholars reach their own conclusions about
the object.
Parallels for the fibula are known from sites in Anatolia and elsewhere.
The main features that characterize the fibula, aside from the precious
material and decorative moldings, are the removable bar or lock-plate and
the nonfunctional spring and double-pin mechanism that is shielded by the
latter. Two types of fibulae that make use of a lock-plate are presently known:
those with a fixed, permanent bar that joins the two ends of the arc; and
those with a removable decorated bar, constructed in the very same manner
as the one found on our gold fibula (Figs. 6, 7). In both types double-pins
rather than the far more common single pin were customary, and the nature
or shape of the fibulas arc and decoration was varied.3
Fibulae of the first type have been excavated at Olympia, the Argive
Heraeum, and Sparta in Greece; at Ine, Alishar, and Ephesus in Anatolia;
and on Samos. Some of these fibulae appear to be Phrygian, others are
apparently copies and not Phrygian, and they all date to the late eighth and
seventh centuries.4
Fibulae of the second type are known from the Phrygian sites of Gor-
dion (Figs. 6, 7) and Ankara in contexts dating them to the late eighth cen-
tury bc; the eighth century fibula carved on the well-known Ivriz rock relief
depicting King Urpallu may also be of this type.5 Other isolated examples
were excavated at Olympia and Samos, and at least one is in the Ankara
Museum.6
3 I summarize here information available in more detail in Phrygian Fibulae, pp. 15, 17f.,
20 f., 22, 24, 39; see also the following footnotes.

4 Phrygian Fibulae, pp. 20 f., 28, n. 5, 29, n. 8, 31, n. 28, 29, 30, 39, Figs. 3540, 41, 42; Fibules,
pp. 217 f., No. XII, 10e from Delphi may be a buckle.
5 Phrygian Fibulae, pp. 15, 17, 19 f., 22, 24, 26, 30, n. 11, 33, n. 51, 39, Figs. 8, 9, 1621, 47, 48,
63, 65.
6 Ibid., pp. 17, 22, 30, n. 14 and 21, 32, n. 35. The examples from Olympia and Ankara do
not have the lock-plate preserved.

Plate III
Fig. 6. Bronze fibula (B900) from Tumulus MM

at Gordion; ca. 725700 bc, Ankara.
Fig. 7. Bronze fibula (B820) from Tumulus MM

at Gordion; ca. 725700 bc, Ankara.
It is the latter type with which we are concerned, for our fibula without
doubt belongs in this category. The detachable lock-plates and other details
of this type of fibula, particularly those from Gordion, Ankara, and Samos,
are constructed in the same manner as noted on the gold fibula: two tubes
to hold the double-pins, a recessed setback to inclose the catch, a catch that
fits under the upper tube leaving the lower one free, a nonfunctional spring
and a decorated spool set below the spring that fits into a forked cut-out on
the plate.
The gold fibula has crisp, sharp moldings and catch, parallels for which
are easily found in eighth century Gordion and elsewhere.7 It was character-
istic of the eighth and early seventh century bronze fibula types with extra
moldings on the arc (Types XII, 13, 13a, 14, and 14a) to exhibit finer workman-
ship than was customary for the later examples. However, several fibulae
with the same neat sharp workmanship do occur on the City Mound at Gor-
dion in the sixth and later centuries.8
On the basis, therefore, of parallels in typology, and because of the fine
workmanship of the arc, moldings, and catch, it would be possible to con-
clude that the gold fibula was made in Phrygia in the eighth century or the
first decade of the seventh century bc. Nevertheless, certain questions that
arise from a style analysis seem to suggest that such a conclusion cannot be
the only viable one.
One of the questions concerns the gold twisted wire tassel. No fibulae
of a Phrygian type known to me have a tassel of any sort. It is indeed
possible that tassels of a perishable material were used and have not been
recovered or recognized in excavation, but, of course, this is not verifiable.9
We may conclude only that there is at present no evidence that tassels were
characteristic of eighth century fibulae.
Twisted gold wire chains like the example on our fibula seem to have
been made over a relatively long period of time. One finds good parallels
in Iran, Anatolia, and the West dating to the second and first millennia bc
and continuing down into the Hellenistic and Roman periods;10 therefore,
the chain by itself cannot help us chronologically.
7 Ibid., pp. 21 ff., Figs 47, 48, 6365; see also Fibules, pp. 220ff., especially Figs. 249, 252,
255.
8 Phrygian Fibulae, pp. 23, 25. See also my discussion of the finds from Tumulus S at
1
Gordion on pp. 4 f., 10, n. 22 and 23, and Figs. 49, 50, 52, 53, 6771.
9 In 1968, right after I first saw our gold fibula, I examined two bronze fibulae from
Tumulus MM at Gordion that were travelling with The Art Treasures of Turkey Exhibition.
One was a Type XII, 7, the other a Type XII, 14 example and both had double pins and lock-
plates. In each example I noticed a neat hole directly behind the spool below the spring:
these holes could have held perishable tassels that are now missing; see the catalogue for the
exhibition, Smithsonian Institution, Washington D.C., 1966, Fig. 102.
10 E. Negahban, A Preliminary Report on the Marlik Excavation (Tehran, 1964), p. 57, Fig. 127;
R. Ghirshman, The Art of Ancient Iran (New York, 1964), p. 114, Fig. 151; F.H. Marshall, Cat-
alogue of Jewellery in the British Museum (London, 1911), Pls. IV (550), LII (24422443), LIV
(2606, 2581), LVIII (2720), LXIX (3024); P. Jacobsthal, Greek Pins (Oxford, 1956), Nos. 643, 647;
Another question concerns the wire scroll motif. I do not know of any
such motif, or related one, that may be presented as occurring on the lock-
plate of any excavated fibulae.11 Those lock-plates known to me are deco-
rated with cut-outs or lattice patterns, or with studs that pierce the plate.12
The lock-plate on the fibula shown in Fig. 7 is decorated with a separate cir-
cular diaper pattern apparently soldered into place; and, indeed, if solder
was employed here, this feature would be a parallel in technique, if not in
motif, to the scroll decoration on our fibula. Scroll patterns on terracotta
architectural plaques are well known in seventh or sixth century contexts
in Anatolia. They occur at Gordion on the City Mound in post-destruction
levels, and at Sardis;13 a common source is implied. It is possible that these
post-eighth-century scroll patterns might supply us with a clue that our
fibula is post-eighth century in date.
Hollow-headed hemispheres are soldered around the scrolls on our
fibula, but there are no rivets that pierce the plate as was common on eighth
century fibulae. A good late seventh or sixth century example of the use of
hollow gold hemispheres, apparently soldered into position on a gold plate
background, is the jewelry found at Aydin in western Anatolia, just south of
Sardis.14 Again we have a clue that we are dealing with a post-eighth century
object.
Finally, there is the construction of the fibula itself, the fact that the arc
is a hollow tube, and that the moldings are constructed of several units
and added individually and soldered in position. Again, no parallel for this
technique is forthcoming. All the many hundreds of bronze fibulae from
Anatolia are cast solid together with the moldings. One might, of course,
H. Hoffman, P. Davidson, Greek Gold (Brooklyn Museum, 1965), pp. 36f., Fig. 1; O.M. Dalton,
The Treasure of the Oxus (London, 1964), Pl. XXI, Nos. 151, 152, p. 40; P. Amandry, Orfverie
achmnide, Antike Kunst, I (1958), 14 f., n. 47, Pl. 12, No. 32; P. Demargne, The Birth of Greek
Art (New York, 1964), Fig. 493.
11 See Negahban, op. cit., p. 47, and Figs. 65, 67 for similar flat wire decoration. Related
wire decoration may also be seen on some earrings recently excavated in the Athenian Agora,
E.L. Smithson, The Tomb of a Rich Athenian Lady, Hesperia, 37, 1 (1968), Pl. 31, Fig. 77b, Pl. 32,
Figs. 77a, b. Both examples cited here are early first millennium bc in date.
12 Phrygian Fibulae, pp. 17, 19, Figs. 1922, 47, 48, 63; Fibules, p. 216, Figs. 243, 244.
13 G. and A. Koerte, Gordion, Ergebnisse der Ausgrabungen im Jahre 1900, JdI, Supple-
ment No. 5 (Berlin, 1905) (hereafter cited as Gordion), pp. 155f., Figs. 138, 139, p. 164, Fig. 148;
R.S. Young, Gordion1950, Univ. Mus. Bull., 16, 1 (1951), p. 6, Pl. III; Progress at Gordion, Univ.
Mus. Bull., 17, 4 (1953), p. 15, Fig. 8; T.L. Shear, Sardis, X (Cambridge, 1926), Pl. XI; H.T. Bossert,
Altanatolien (Berlin, 1942), No. 183. The pattern may be Greek in origin.
14 A. Dumont, Notes sur des bijoux d or trouvs en Lydie, BCH, 3 (1879), pp. 129f., Pls. IV,
V; P. Demargne, op. cit., Fig. 494.

suppose that it was economy that suggested using a hollow arc because gold
was expensive. It can only be stated, however, that the few gold fibulae found
to date at Gordion (Fig. 8), all from the eighth century, are types with plain
arcs and all are cast solid.15 One site in Asia Minor that has yielded gold or
electrum fibulae is Ephesus, where about seven examples were found in the
Basis and are ca. 600 bc in date.16 One of these fibulae (Hogarths Pl. V, No. 3),
a XII, 14 type, invites our attention as it seems (from the photograph) to
have a hollow arc; the other examples illustrated all seem to have solid arcs.
Whether this fibula actually has a hollow arc is not known to me, as the
text gives no information on the matter; I also cannot discover the present
location of the fibula. Certainly, if the arc is hollow, it would be an important
parallel for our Type XII, 14 fibula.
Collecting all the above observations together and attempting to inter-
pret them culturally and chronologically suggests to me the following con-
clusion: the fibula is either an eighth or early seventh century bc Phrygian
object of a hitherto unknown technique for that period, or it is a later copy
of a Phrygian product. If the latter possibility proves correct, I would sug-
gest that it is Lydian and that it was made sometime in the sixth century bc.
In other words, the fibula could have been made any time between the late
eighth and the sixth century bc. My inclination, however, is to think that
the latter conclusion, that the fibula is sixth century, and probably Lydian,
is actually the correct one.
The gold fibula is not an isolated example of an object that illustrates
the problem of sorting out eighth from sixth century bc or Phrygian from
Lydian products. For what emerges from a study of the artifacts excavated
in post-Kimmerian Gordion is that several eighth-century forms and motifs
continue apparently unchanged into archaic times, the sixth century. And
when one examines the small but growing number of objects becoming
available from Lydia, one notices in some examples a dependence on Phry-
gian culture. At one time I was under the impression that the late seventh
and sixth century Lydian parallels to Phrygian objects reflected a knowl-
edge of eighth century culture there, but I now believe that the evidence
suggests Lydia borrowed objects and ideas from contemporary Phrygian cul-
ture.
15 Gold jewelry was not common at Gordion until the sixth century, Phrygian Fibulae, p. 9,
n. 8; p. 44 for the gold fibulae found at Gordion and Ephesus.

16 D.G. Hogarth, Excavations at Ephesus (London, 1908), p. 98, Pl. V, Nos. 15; Phrygian
Fibulae, pp. 14, 28, n, 5, 44.

It is possible in this brief study to give only some examples to support

the above statement mainly because not much material has been published.
Also, because it is more important at present, I think, to suggest the way for
future areas of study for those more knowledgeable in the subject. Collec-
tively, the evidence does allow one to come to two meaningful conclusions: a
study of the fibulae, metal bowls, pottery, and tumulus construction at Gor-
dion suggests that there was a limited but nevertheless recognizable cultural
continuity there from the eighth through the sixth centuries bc; and at the
same time, it is possible to document Lydian knowledge of each of these
items.
Fibulae were quite common in the eighth century tombs at Gordion
where hundreds have been excavated. In the seventh and sixth centuries
the number drops, and only a few tombs of this period contained fibulae.
But all levels of the City Mound, from the sixth to the second centuries bc,
contained fibulae, especially Types XII, 13 and 14.17 This demonstrates that
fibulae were manufactured (or at least used) at Gordion for centuries after
the Kimmerian destruction. At Bogazky there is also evidence to prove this
point, and at Karalar a tomb of the first century bc yielded a Type XII, 14
fibula.18
It is understood, then, that with respect to certain fibulae, in particular
Types XII 13 and 14, a problem may arise about the actual date of a given
specimen. To be sure, as with all the objects discussed in this study, the
problem is more acute when the fibula is studied out of context from the
objects excavated with it (such as a fibula purchased on the antiquities
market!); if one has other objects from the same tomb or level, the analysis
is less difficult.19
There is no evidence at present to show that Lydia in the eighth century
used Phrygian fibulae,20 but two Phrygian-like fibulae of Type XII, 14 were
17 Phrygian Fibulae, passim, pp. 78 f., Appendix A and B. Problems of dating Types XII, 13
and 14 fibulae are discussed on pp. 3, 22 f., 39 f.

18 Ibid., pp. 25, 28 f., n. 52, 33, n. 45. For a discussion of the difficult chronology of Midas
City and Alishar see pp. 29. n. 8, 30, n. 14.

19 Ibid., pp. 3, 39 f. Note that differences of opinion do indeed exist concerning the dating
of Anatolian tombs, cf. Phrygian Fibulae, pp. 17, 26, 30, n. 11 with comments by E. Akurgal,
Phrygische Kunst (Ankara, 1955), pp. 83, 130; Akurgal dates all the Ankara tumuli to the
seventh century bc. Makridys Tumulus II, apparently the third tumulus which he dug, is
seventh or early sixth century in date, see n. 25.
20 G.M.A. Hanfmann, The Ninth Campaign at Sardis (1966), BASOR, 186 (1967), p. 29,
Fig. 14, shows a fibula from Sardis dated to the Protogeometric period. On p. 36, n. 7 a parallel
from Skyros is cited; better parallels may be seen in Fibules, Type IV, 11, examples of which
found in a Lydian level at Sardis dated to the late seventhearly sixth cen-
tury bc; whether these were products of local Lydian workshops or were
imported from Phrygia is not known.21 In a context ca. 600bc at Ephesus we
have seen that Phrygian-type fibulae of gold and electrum were found. Here,
too, it is not possible to conclude whether these fibulae were locally made or
whether they were imported from Lydia or Phrygia.22 And other sites in the
area, such as Bayrakli, Larisa, Chios, and Lesbos, have yielded Phrygian-like
fibulae from the seventh to the fifth centuries which thus demonstrate that
these types were locally manufactured during those periods.23
Bronze bowls, with or without omphaloi, and often decorated with petals,
or with spools and elaborate handles, are found in all eighth century tombs
at Gordion and Ankara.24 At both sites, evidence exists that these bowls
were still being manufactured in the seventh and sixth centuries (or at
least were still available as heirlooms during this period). Tumulus J, of
the seventh century, and Tumuli F, M, and E, of the sixth century, all at
Gordion, contained in their tomb deposits various products among which
were bronze omphalos and other bronze bowl types. And one of the tumuli
excavated by Makridi at Ankara, of seventh or early sixth century date,
contained bronze bowl fragments.25
come from several of the Aegean islands and from Ephesus; see also N. Firatlis report in
Annual of the Archaeological Museums of Istanbul (Istanbul, 1958), pp. 75f., Fig. 13, Nos. 14,
from an area near Izmir.
21 Phrygian Fibulae, pp. 25, 44, Figs. 75, 76, p. 33, n. 45.
22 Ibid., p. 44, for a discussion of some of these fibulae as Lydian objects.
23 Ibid., pp. 31 f., n. 31, 35, 36, 37, 39 f. Belt buckles in the form of fibula arcs occur in
the eighth century and continue through the seventh, J. Boardman, Ionian Bronze Belts,
Anatolia, 6 (19611962), 179 ff.
24 Gordion, p. 73, Nos. 6072, Figs. 53, 54; p. 74, Nos. 7380; R.S. Young, Gordion 1956:
Preliminary Report, AJA, 61, 4 (1957), p. 328, Pl. 93, Fig. 31; The Gordion Campaign of 1957:
Preliminary Report, AJA, 62, 2 (1958), p. 150; The Gordion Campaign of 1959: Preliminary
Report, AJA, 64, 3 (1960), p. 230; T. zg, M. Akok, Die Ausgrabungen an zwei Tumuli auf
dem Mausoleumshgel bei Ankara, Belleten, XI, 41 (1947), pp. 5785; H. Zubeyr, Ankara Gazi
Orman Fidanliginda Buhman Eserler, TTAED, I (1933), 10f., Nos. 13, p. 12, No. 5; see n. 18 for
Akurgals late dating of these tumuli.
25 Gordion: Univ. Mus. Bull., 17, 4 (1953), pp. 30 f., 34 f.; Univ. Mus. Bull., 16, 1 (1951), pp. 17f.;
E. Kohler, E.K. Ralph, C-14 Dates for Sites in the Mediterranean, AJA, 65, 4 (1961), p. 362;
Tumulus M is not yet published. Ankara: M. Schede in AA (1930), cols. 481f.; T. zg,
Untersuchungen ber archologische Funde aus Anatolien, Belleten, 10, 40 (1946), p. 619;
E. Akurgal, Phrygische Kunst, pp. 84, 130 for dating. There seems to be some disagreement
about the provenience of a bronze vessel fragment: zg, op. cit., p. 617, fig. 33, calls it an
object from Makridys second tumulus, whereas Akurgal in Bayrakli (Ankara, 1950), p. 89, says
it came from the third tumulus excavated by Makridy.
In Lydian territory there is rare but significant evidence that bronze bowls
of Phrygian types were in use. A plain bronze bowl with Phrygian-type
handles and spools was found in a grave at Manisa, near Sardis; its date is
stated to be seventh or sixth century bc.26
A small bronze omphalos bowl with a raised central groove was found
at Ephesus in a poor stratigraphical area.27 Its size is smaller than most
Phrygian bowls and it is possible that it was a local product and not imported
from Phrygia.
One of the best pieces of evidence that Phrygian-type bowls were in use
in the sixth century bc in the west is the apparently metal bowl carried by
the well-known ivory lady found at Ephesus.28 This bowl could certainly be
accepted as an eighth century product of Phrygia. Unfortunately, we are not
in a position to establish whether the bowl carried by the lady was an import
from Phrygia, or Lydia, or whether it was made in the Greek East.
One more example of a sixth century date for a Phrygian-type bowl is
a petalled specimen found at Trebenishte in Yugoslavia.29 There can be no
doubt about the date of the grave in which it was found, but we are not able
to make a specific statement about the bowls original home or whether it
was an heirloom.
With the omphalos bowls one is, therefore, faced with a problem parallel
to the one discussed above concerning our gold fibula. Thus, if a Phrygian-
type bowl is encountered on the antiquities market isolated from its original
context, it would be difficult to decide whether it is Phrygian or Lydian,
or whether it is eighth century, or later. A silver petaled omphalos bowl
allegedly from Anatolia now in the collection of the Metropolitan Museum
of Art (Pl. IV, Fig. 9), and another similar example in the Birmingham City
Museum are two such examples.30 Another silver omphalos bowl, plain and
26 Akurgal, Phrygische Kunst, p. 81, Pl. 60a, dates it to the sixth century: A.K. Knudsen,
From a Sardis Tomb: A Lydian Pottery Imitation of a Phrygian Metal Bowl, Berytus, XVI
(1964), 63 dates it to the seventh century. Compare the Manisa bowl to examples from the
Baumschule tumulus at Ankara, Phrygische Kunst, Pls. 58a b, 59a, b, and Gordion, R.S. Young,
Gordion on the Royal Road, PAPS, 107, 4 (1963), p. 362, Fig. 19.
27 Hogarth, op. cit., p. 152, Pl. XV, 13.
28 Ibid., Pls. XXI, XXII; Knudsen, op. cit., p. 64, n. 11; Akurgal, Phrygische Kunst, pp. 81f. I
agree with Akurgal and Knudsen that the bowl reflects Phrygian rather than Greek influ-
ences.
29 B. Filow, Die archaische Nekropole von Trebenishte (Berlin, 1927), pp. 75f., Fig. 93, 1;
P. Jacobsthal, Greek Pins, pp. 203, 208.

30 Archaeological Reports for 19641965, The Society for the Promotion of Helennic Studies
(1965), p. 63, Fig. 1.

with grooves around the omphalos, now in a private collection and said
to come from Anatolia, fits into this category (Pl. IV, Fig. 10). The size of
these three bowls conforms to that of the bronze examples excavated at
Gordion and Ankara, and in typology they appear to be eighth century
Phrygian products. Yet, in Anatolia, aside from a silver omphalos bowl found
at Bogazky of uncertain date, all of the many bowls excavated at Gordion
and Ankara are made of bronze (two bronze examples from Tumulus 1 at
the Mausoleum Hill in Ankara are said to be gold plated).31 And none of the
petaled bronze bowls found at Gordion has the incised lozenge decoration
around the omphalos that exists on the silver bowl shown in Fig. 9. It is,
therefore, temptingbut certainly premature at presentto conclude that
these silver bowls might be Lydian and sixth century bc in date. Certainly
one must be prepared to entertain such a possibility. What is needed, of
course, is a detailed study of the techniques and styles of the many excavated
bowls from the Phrygian tombs before one generalizes about unexcavated
material appearing on the market.32
Imitations of Phrygian metal bowls made in terracotta were manufac-
tured in the eighth century at Gordion and Bogazky.33 This practice contin-
ued at Gordion into the later centuries, for pottery vessels imitating metal
shapes of typical eighth century types have been excavated on the City
Mound in the archaic levels.34 One late example from Gordion, a spouted
pitcher in a local black-polished ware, a masterpiece of craftsmanship, imi-
tates in all detailsbolsters, studs, and separate handlea metal prototype
(Pl. V, Fig. 11). The excavator made these pertinent remarks: the Phry-
gians must have kept right on producing elaborate vessels of metal down
into the fifth century, for the shape of our jug is typically Phrygian, known at
31 Bogazky: H. Otto, Die Funde von Bykkale, MDOG, 78 (1940), 48f., Fig. 9, 1, aus der
ersten Schicht; the bowl does not appear to me to be a typical eighth century type and seems
to be later. T. zg, M. Akok, op. cit., Pl. X, Fig. 18 for one of the two allegedly gold-plated
examples. Also see note 36 below for a comment about silver vessels in the sixth century bc.
32 See for example M.-L. Buhl, Skatte fra Det Gamle Persien, Nationalmuseet (Copenhagen,
1968), pp. 86 f., No. 199 with figure on p. 87: a petaled bronze omphalos. The caption says
the bowl is Greek, was found in Iran (presumably it was bought from a dealer), and is sixth
century in date. Judging from the photograph the bowl appears to be eighth century Phrygian!
33 Gordion, pp. 66 f., Fig. 42; R.S. Young, AJA, 61, 4 (1957), pp. 328f., Pl. 95, Fig. 33; T. Beran,
Eine Kultsttte phrygischer Zeit in Bogazky, MDOG, 94 (1963), 45ff., Figs 13, 14; Knudsen,
op. cit., pp. 65 ff., discusses this object as well as the use of bronze handles associated with
wooden bowls.
34 Gordion, pp. 196 f., Fig. 195; Knudsen, op. cit., p. 67, n. 22, and infra. Cf. Eva-Marie Bossert,
Funde nachhethitischer Zeit, MDOG, 89 (1957), 67 f., Fig. 53.

Plate IV
Fig. 8. Gold and electrum fibula from the destroyed

Phrygian level at Gordion (top J131, bottom J130);
late eighth-early seventh century bc, Ankara.
Fig. 9. Silver petaled Fig. 10. Silver omphalos;

omphalos; anonymous private collection, U.S.A.
gift to the Metropolitan
Museum of Art, 66.116.
Gordion as early as the eighth century, and the potter must have been using
as his model a nearly contemporary metal vessel of local manufacture.35
Another example from Gordion, an omphalos fragment in black polished
ware, was found in a reused Phrygian house and could belong to either
Phrygian or Persian times; the excavator also remarks: the omphalos
fragment is like that shown by many of the bronze bowls from the Phrygian
tombs.36
One should call attention in this context to the glass omphalos found in
the eighth century tomb of Tumulus P at Gordion. It, too, imitates a standard
metal prototype, similar to those excavated at contemporary Gordion, and
it neatly demonstrates that imitations of metal vessels were not confined to
terracotta.37
We may conclude from these brief comments that, albeit on a limited
scale, Phrygian craftsmen in the sixth and fifth centuries bc manufactured
pottery of the same types as did their eighth century ancestors.
Turning our investigation now to Lydia, we see a similar situation obtain-
ing. Lydian craftsmen of the sixth century imitate Phrygian metal vessels
in pottery and at the same time they also copy Phrygian pottery vessels.
Pottery of these types, found both in metal and pottery at Gordion and
elsewhere, have been excavated at Sardis: side-spouted pitchers with strain-
ers,38 bowls with spool handles and bolsters (Pl. V, Fig. 12), and horizontal
35 R.S. Young, The 1963 Campaign at Gordion, AJA, 68, 3 (1964), pp. 282f., Pl. 84, Fig. 12.
I would like to thank Professor Young for allowing me to republish the vessel here for the
convenience of the reader. We would like to know more about the fabric of this vessel: is the
fabric different from that of eighth century examples of black polished wares?
36 R.S. Young, The Gordion Campaign of 1967, AJA, 72, 2 (1968), p. 235, Pl. 76, Fig. 15. See
also R.S. Young, The 1961 Campaign at Gordion, AJA, 66, 2 (1962), pp. 154f., Pl. 41, 1, 2 for
a silver Achaemenid vessel and its counterpart in black polished ware found in the archaic
level. Incidentally, this silver example informs us that in the sixth century at Gordion silver
vessels were in use.
37 R.S. Young in AJA, 61, 4 (1957), p. 328, Pl. 94, Fig. 32. Young considers the bowl to be an
import but cf. Axel von Saldern in Glass Finds at Gordion, Journal of Glass Studies, I (1959),
22, 25 ff., Figs. 1, 2; Von Saldern accepts the possibility that the bowl was a local product but
prefers to see it as an import.
38 Gordion: AJA, 64, 3 (1960), Pl. 56, Figs. 6, 8 (the latter is bronze); AJA, 68, 3 (1964). Pl. 84,
Fig. 12 (here Fig. 11); Gordion, pp. 57 f., Figs. 2123; pp. 62ff., Figs. 28, 29, 3134; Akurgal, Phry-
gische Kunst, Pls. 11, 12, 14b. Cilicia: H. Goldman, Excavations at Gzl Kule, Tarsus (Princeton,
1963), Pl. 88, 1242. Alishar: H.H. von der Osten, The Alishar Huyuk, Seasons 193032, II (Chicago,
1937), OIP, XXIX, 365 f., Figs. 414415. Bogzaky: K. Bittel, H.G. Gterbock, Bogzaky (Berlin,
1935), p. 57, Pl. 13a, b (cf. Fig. 11 in this study). Sardis: Phrygian Fibulae, p. 44, Fig. 104, p. 47, n. 30
(= Andrew Oliver, Jr., Lydia, BMMA, XXV, 5 (1968), p. 199, Fig. 8 upper left): see also Phrygian
Fibulae, pp. 66, 74 f., n. 45.
Plate V
Fig. 11. Black polished vessel (P2805) from Gordion,

City Mound archaic level, sixth-fifth century bc, Ankara.
Fig. 12. Lydian pottery vessel (P1620) from Tomb 23a

at Sardis; the Metropolitan Museum of Art, 14.30.20.
supports,39 and a side-spouted vessel with a cover attached to and sliding

along the handle.40 The majority of the Phrygian vessels of these types occur
in the eighth century bc, but some are known from the seventh and sixth
centuries bc, as discussed above. Therefore, we may conclude that Lydians
used contemporary imported Phrygian shapes as models. A scholar who has
recently studied one variety of pottery imitation excavated at Sardis, a bowl
with spool handles and bolsters, remarks about a particular bowl that it is
a close imitation of an imported contemporary bronze Phrygian bowl 41
The Lydian examples of Phrygian shapes usually have a splayed foot on the
pitchers as well as distinctive painted decoration, so that it is usually not
difficult to differentiate Lydian from Phrygian vessels.42
At Gordion in post-Kimmerian times Lydian material was imported
showing that trade moved in both directions. Lydions, the very characteris-
tic Lydian vessel of the sixth century, were common on the Kchk Hyk,
a small mound adjacent to thc City Mound, and at least two tumuli, A and
C, had Lydian pottery in their tombs.43
The presence of the star and scroll pattern on terra cotta architectural
plaques at sixth century Gordion and Sardis, already referred to above, is a
further indication of contact between the two cities.
One further item which is significant in a discussion of Phrygian and
Lydian relationships is the similarity of their burial techniques. It was a com-
mon practice for Phrygian tombs to be placed away from the center of an
39 Gordion: Gordion, pp. 72 ff., Figs 5054; AJA, 62, 2 (1958), Pl. 25, Figs. 15, 21; Pl. 27, Figs. 22,
24; R.S. Young, The Gordion Tomb, Expedition, I, 1 (1958); AJA, 61, 4 (1957), Pl. 93, Fig. 31.
Ankara: Akurgal, Phygische Kunst, Pls. 5759. Sardis: Oliver, op. cit., p. 199, Fig. 8 lower right;
H.C. Crosby, Sardis, I (Leyden, 1922), 119, Fig. 126 center (= Fig. 12 here); Knudsen, op. cit., Pls.
VIII, IX.
40 The Gordion metal example from the eighth century Tumulus MM has not yet been
published. Sardis: Phrygian Fibulae, Fig. 103 (= Oliver, loc. cit., upper right). See Von der Osten,
op. cit., p. 150, Fig. 191, for a Hittite version from Alishar.
41 Knudsen, op. cit., p. 67.
42 Compare the low foot or lack of a foot on the side-spouted vessels from eighth century
Gordion referred to in note 36 to the foot on the sixth century examples from Gordion (Fig. 11
here) and Sardis. The example from Bogazkoy has a low foot, the one from Tarsus has no
foot. G.F. Swift, Jr., The Ninth Campaign at Sardis (1966), BASOR, 186 (1967), 34, refers to
p. 24, Fig. 8 as a Lydian Geometrio pottery fragment that imitates Greek Geometric styles. It
seems to me that a Phrygian prototype could actually have served as a model, e.g. Gordian,
pp. 58 ff., Figs. 22, 25, 26. For a discussion of the relationships between Greek and Phrygian
Geometric pottery see Akurgal, Phrygische Kunst, pp. 33ff., 121, 126 and supporting comments
by this writer in Phrygian Fibulae, pp. 60 ff.
43 Phrygian Fibulae, pp. 6 f., 8, 10, n. 36 and 37, 44; R.S. Young, Making History at Gordion.
Archaeology, 6, 3 (1953), pp. 164 f., Figs. 8, 9.

overlying tumulus. This practice began in the eighth century and continued
into the seventh and sixth centuries. The Lydians also built tumuli over their
tombs, and these tombs were usually placed off center from the middle
of the tumulus. It seems almost certain that the idea was borrowed from
the Phrygians. The very same type of tumulus constructon with the grave
off center was known in northwest Iran and in Cyprus during the first
millennium bc, and the idea may have spread to these areas from Phrygia.44
Future research will help us solve some of the issues and problems raised
in this study. We know very little about stylistic differences between eighth
and sixth centuries bc Phrygian pottery, especially of the black-polished
varieties, because no comprehensive study of the Gordion or Bogzaky
pottery has yet appeared. Likewise, we have no published analysis of the
bronze bowls from the eighth century tombs comparing them to the few
sixth century examples found in the later tombs. We would like to know
something about the types of metals and techniques used to manufacture
them, as well as a stylistic discussion. From the Lydian side we would also
want a study of Lydian pottery so that we may better be able to discuss
the extent of Phrygian influences. And knowing this, we would be able to
compare the relative values of both Greek and Phrygian (Near Eastern)
influences on Lydian culture.
44 The evidence is summarized in Oscar White Muscarella, The Tumuli at S Girdan: A
Preliminary Report, Met. Mus. of Art Journ., II (1969), 7 ff.

chapter twenty-seven
FIBULAE AND CHRONOLOGY, MARLIK AND ASSUR*
Abstract: In two recent publications of material from tombs excavated at two Near
Eastern sites, Marlik Tepe in NW Iran and Assur in Assyria, the occurrence of
fibulae among the grave goods and the concomitant chronological conclusions to
be obtained from their presence have been either ignored or distorted. In this paper
it is argued that a correct analysis of the chronology of the advent of fibulae in
the Near East leads to significant results, results other than those perceived by the
scholars responsible for the publications.
Still, it is an error to argue in front of your
data, you find yourself impossibly twisting
them round to fit your theories.
Sherlock Holmes
In their discussion of, respectively, Iranian and Urartian fibulae (the term
for ancient safety pins), Roman Ghirshman and Baki gn both thought
it appropriate to quote Solomon Reinachs comment made in 1896, that
the fibula is un principe de classification chronologique et que la fibule
devait tre considr comme ce fossile directeur de l antiquaire .1 Rei-
nachs old statement remains cogent and meaningful, all the more so as
subsequent research has confirmed his conclusion in general terms and,
with more refined determination of stratified finds, has allowed scholars to
consider the fibula to be one of the most important chronological indicators
available in archaeological research. This situation obtains for Europe and
the Aegean, where fibulae first appear in the late 2nd millennium bc, as well
as in the Near East. One might say the fibula, wherever found in an Old World
context, is a precise chronological indicator, a Leitfossil, at least in regard to
post and ante quem dating.
Speaking only with concern for the Near East (fig. 1), it may be asserted
with some certainty that except for a relatively small number of fibulae
* This article originally appeared as Fibulae and Chronology, Marlik and Assur, Journal
of Field Archaeology 11, no. 4 (1984): 413419.

1 Roman Ghirshman, La Fibule en Iran, IranAntiq 4:2 (1964) 90; Baki gn, Urartische
Fibeln, (AMI Ergnzungsband 6: Berlin 1979) 178.

804 chapter twenty-seven
Figure 1. Map of the Near East.
known as East Mediterranean types that occur in western Palestine and

western Syria, and which are mainly of Aegean and Cypriote origin or deriva-
tion, the Near Eastern fibula types are phenomena of the late 8th century bc
and later.2 Painstaking research and examination of the many sites that
2 David Stronach, The Development of the Fibula in the Near East, Iraq 21:2 (1959)
181206. For the East Mediterranean fibulae, see 182185, fig. 1, Map 2. For the Near Eastern
examples, see 185206, especially 185 and 193 for chronology. The dating of the East Mediter-
ranean Type II, with round or arched bow and with mouldings or beads at the terminals,
is somewhat problematic. It may have come into existence before the late 8th century but
by no means with certainty. Part of the problem concerns the occurrence of Type II fibu-
lae at Megiddo and the chronology of the excavated levels there. The excavators, R.S. Lamon
and G.M. Shipton, Megiddo I (University of Chicago Press: 1939) xxvii, give a chart with a
tentative chronology of Levels I, II, III, and IV, but warn that it must be used with utmost
fibulae and chronology, marlik and assur 805
yielded fibulae in Syria, Anatolia, Urartu, Assyria, and Iran has charted the
chronological position of these artifacts. Initially two scholars, Ch. Blinken-
berg and D. Stronach, documented the occurrences and stratigraphy of hun-
dreds of Near Eastern fibulae and classified them into various types.3 They
exist in numerous varieties and forms. Some are semicircular or arched,
either with plain or beaded arms (Stronachs Types I and II). Others, by far
the most common and ubiquitous, are triangular in shape (often referred
to as knee or elbow fibulae), the great majority decorated on the arms with
discs, ribs, and beads (Stronachs Types III and IV); the catch is flat, bent to
receive the pin, and in a large number of examples is cut in the form of a
human hand. Exactly when the hand catch first appears is uncertain, but it
can be documented nowhere before the late 8th century, nor does it appear
earlier on Cypriote and East Mediterranean fibulae.
More recent scholarship has but reinforced and further clarified the
fibula chronology proposed by Stronach for Syria and Mesopotamia, and
the interested student is invited to study the evidence independently.4
reservations, hardly a basis for discussions of absolute chronology for the fibulae. On pls. 78
and 79 are examples of East Mediterranean and Near Eastern fibulae from Levels I, II, and III,
which they date to the 8th century and later; only one fibula, pl. 78:20 (= Stronachs Type II, 4,
fig. 6:4), derives from Level IV filling, i.e., presumably pre-8th century (1000800 bc). In any
event, it derives from the fill, not from a locus in Level IV. In G. Loud, Megiddo II (University
of Chicago Press: 1948) only one fibula, pI. 223:79, is listed as from Level IV; it is isolated in
the text and presented with no discussion. I therefore suggest that Stronach (op. cit., in this
note, 191) misstates the evidence that his fig. 6:4 emerges about 900 bc. There is no evidence
that compels one to accept as fact that Type II fibulae are earlier at Megiddo than the typical
Near Eastern examples; they coexist there together.
3 Ch. Blinkenberg, Fibules grecques et orientales (Copenhagen 1926), Type XII, 2425, 204
225; Type XIII, 236238, 239 247. This monograph is still the handbook and the initial study
for all Fibelforschers; Stronach, op. cit. (in note 2).
4 M.V. Seton-Williams, Preliminary Report on the Excavations at Tell Rifa"at, Iraq 23:1
(1961) 86, pl. XLI, 11, 12; Judy Birmingham, The Development of the Fibula in Cyprus and the
Levant, PalExplorQuar 95 (1963) 80112, 91, 95 (but with imprecise chronologies and strati-
graphical references); D.E. McGown et al., Nippur I (University of Chicago Press: 1967) 6971,
99, pl. 151:1921; P.R.S. Moorey, Cemeteries of the First Millennium bc at Deve Hyk (BAR Inter-
national Series 87: 1980) 8586; Arndt Haller, Die Grber und Grfte von Assur (Verlag Gebr.
Mann: Berlin 1954) 12161 (for a total of 99 fibulae from 65 graves and tombs, all neo-Assyrian).
This last publication is impossible to use with regard to a diachronic understanding of fibula
types recovered in the earlier and later neo-Assyrian burials. Haller makes it quite clear on
pages 5 and 12 that his division of the burials into a Frhe Zeit 1100824 v. Chr. and a Spte
Zeit 824612 v. Chr. is not firm and absolute. The division is based in part on a few inscrip-
tions and on an interpretation of the finds, for which he significantly states (page 5) In vielen
Fllen werden in der Datierungsfrage Zweifel entstehen . Of the objects used to make the
division, he further notes that they are mehr oder weniger charakteristisch. Haller posits
that Armfibeln in runder Form occur in the earlier burials, triangular ones in the later,
but of the 99 fibulae mentioned he illustrates only one, a triangular example from Gruft 30
Furthermore, the most recent studies targeted on the distribution and chro-
nology of the fibulae from Anatolia,5 Iran,6 and Urartu7 have also reinforced
the previously held conclusions. These studies, based surtout sur les
tmoignages tirs des circonstances de trouvaille, to quote Blinkenbergs
criterion for la determination chronologique des fibules,8 have indicated
that no fibula was manufactured or in use in any of these areas before the
late 8th century bc. Before that time straight pins served to fasten clothing.
(pl. 22), which he dates to the earlier part of his putative later period, late 9th century bc (110;
see infra for a discussion of this fibula). Hallers terminology in his description of the fibulae
is unclear and inconsistent: Fibula, Bronzefibula, Armfibula, eckige Armfibula; in only
three cases does he refer specifically to a rundgebogene Fibula (Grab 681), Arm fibula in
rundlicher Form (Grab 758), gerundete Armfibula (Grab 809), none of which is illustrated.
Grave 758 had only anklets along with the fibula as burial contents. All we can say for the
Assur fibulae is that they occur only in the neo-Assyrian period, and suggest that none has
been demonstrated to occur before the late 8th century bc.
5 Oscar White Muscarella, Phrygian Fibulae from Gordion (Bernard Quaritch Ltd: Lon-
don 1967); R.M. Boehmer, Die Kleinfunde von Bogazky (Gebr. Mann Verlag: Berlin 1972)
4666; idem, Phrygische Prunkgewnder des 8. Jahrhunderts v. Chr. AA (1973) 151, 166, 172.
Recent discussions of Phrygian fibulae imported into the Greek spheres reinforce the Ana-
tolian chronology: E. Sapouna-Sakellarakis, Die Fibeln der griechische Inseln. Prhistorische
Bronzefunde XIV, Band 41 (Munich 1978) 120129; Klaus Killian, Fibeln in Thessalien von der
mykenischen bis zur archaischen Zeit. Prhistorische Bronzefunde XIV, Band 2 (Munich 1975)
151154. An earlier date for the advent of fibulae at Gordion is suggested by R.S. Young, Three
Great Early Tumuli (University Museum: Philadelphia 1981) 9, based on his belief that Tumu-
lus W, which contained fibulae, was to be dated to the end of the 9th century bc or slightly
later. For a criticism of this chronology see Oscar White Muscarella, King Midas Tumulus
at Gordion, The Quarterly Review of Archaeology (December 1982) 89.
6 Roman Ghirshman, op. cit. (in note 1) 92; Oscar White Muscarella, A Fibula from
Hasanlu, AJA 69:3 (1965) 233240; P.R.S. Moorey, A Catalogue of the Persian Bronzes in the
Ashmolean Museum (Oxford University Press: Oxford 1971) 124, 127; Louis vanden Berghe,
Les Fibules provenant des Fouilles au Pusht-i-Kuh, Luristan, IranAntiq 13 (1978) 3574 (a
comprehensive and most valuable summary of all the fibulae found to date in Iran; for
chronology see pages 49, 67, 69; add W. Kleiss et al., Bastam I [Gebr. Mann Verlag: Berlin
1979] 171:7, 9, 172:3; and Marlik, see infra). The evidence of the excavations at Hasanlu in
NW Iran is telling: no fibulae were recovered in the settlement or cemetery of Period IV,
which terminated ca. 800 bc, but they occur in Periods III and II, the former, Iron Age III,
commencing in the late 8th century bc or later. Note that all the spring fibulae excavated
in Iran are indistinguishable from those excavated in Syria and Assyria. A great number
of stray, Near Eastern type fibulae published by many scholars (including Ghirshman and
Stronach, and many others too numerous to cite) as deriving from Iran in fact derive from
the antiquities market, i.e., not from Iran, archaeologically speaking. For one special group of
misattributed fibulae see Oscar White Muscarella, Five Additions to the Norbert Schimmel
Collection, APA 7/8 (19761977) 316, no. 3.
7 Muscarella, 1965 op. cit. (in note 6) 235237; for a comprehensive review of fibulae from
Urartu see gn, op. cit. (in note 1) 178188. On page 178 gn dates the advent of fibulae in
Urartu vor dem letzten viertel des VIII. Jahrhunderts nicht.
8 Blinkenberg, op. cit. (in note 3) 26.
In summary, and to the point, after much research and review by a large
number of scholars of all the information available from excavations, the
evidence is certainas certain as is possible in archaeological activity
that the presence of a fibula in a stratified or closed context in the Near East,
at least in areas east of the eastern Mediterranean zone, signifies an ante
quem non date of late 8th century bc or later. And only the occurrence of a
Near Eastern fibula in an unequivocally dated closed context from an earlier
period will alter these conclusions; but I submit that such an anticipated
occurrence seems highly unlikely given the evidence to hand. I know of few
other classes of artifacts aside from pottery that, by all the copy book rules
and concepts of archaeological conclusion formation, can be accepted with
some ease as a reliable fossile directeur, The fibula in the Near East is a
classic, paradigmatic, ante quem non artifact.
The above introductory and background comments are presented to serve

as a framework and milieu for the main purpose of this paper: to call atten-
tion to two recent publications of finds from two separate Near Eastern sites
that, by ignoring or avoiding the corpus of work on fibulae, have missed sig-
nificant chronological and archaeological conclusions. One site is in Assyria,
the other in Iran where indeed the issue is of major importance for the his-
tory of ancient Iran. I will first address the latter case, which concerns the
chronology of the cemetery at Marlik Tepe, situated in the Gohar-Rud valley
in Gilan, just south of the Caspian Sea in NW Iran.
Marlik Tepe was excavated by the Iranian archaeologist Ezat Negahban
in 1961 and 1962 under extraordinary and uncommon difficulties and local
pressures, primarily from antiquities dealers and their supporters. Fifty-
three intact tombs were uncovered, many containing sumptuous gold, silver,
and bronze vessels, gold jewelry, weapons and utensils, bronze and terra-
cotta figurines of humans and animals, seals, and so forth. The first scholarly
report on the site appeared in 1964.9 In a series of subsequent publications
over the years (1965, 1977, 1979, 1981, 1983),10 Negahban has published more
material from the site in the same partial form as in 1964, and in these
instances was concerned solely with specific classes of artifacts. To date,
9 Ezat Negahban, A Preliminary Report on Marlik Excavation (Teheran 1964).

10 Notes on Some Objects from Marlik, JNES 24:4 (1965) 309327; The Seals of Marlik
Tepe, JNES 36:2 (1977) 81102; Pottery and Bronze Human Figurines from Marlik, AMI 12
(1979) 157173; Maceheads from Marlik, AJA 85:4 (1981) 367378; Metal Vessels from Marlik.
Prhistorische Bronzefunde II, Band 3 (Munich 1983).
not a single tomb group with complete inventory and illustrations has been
published (Negahban has informed me that the full publication of all the
tomb groups is forthcoming). In his 1964 and 1977 reports Negahban gave
in the texts only the excavation grid numbers (in Roman numerals) for the
locus of the objects he illustrated (in 1977 in a somewhat confusing manner,
e.g., tomb VI B+, tomb XVIII C, or grid XX F); in 1979, 1981, and 1983
he gave for each object illustrated both a grid (Trench) Roman numeral
and a tomb number (in Arabic numerals), thereby clarifying for the first
time the important distinction between grid number and specific tomb
number.11 More importantly, in the 1979 and 1981 reports a complete plan
was published with each tomb drawn in its respective grid and identified
by number (Tombs 1 to 53). Working through all the tomb and artifact
references, matching and sorting out grids and tombs using the grid-tomb
plan as a guide, and based on what has been published with illustrations,
I have been able to isolate the partial contents of about 31 of the 53 Marlik
tombs. Here I will discuss the contents of only one of the tombs isolated. It
is, as will become evident, one of the most important tombs in the Marlik
cemetery.
In his most recent publication of the Marlik finds (1983) Negahban illus-
trates and discusses 63 metal vessels, 15 of gold, six of silver, and 42 of bronze,
excavated from 32 identified tombs. On pages 8990 we are given a list (but
without previous publication references, except for the vessels) of all the
artifacts found in one of the richest [tombs] in Marlik both in terms of
quantity and quality of the burial gifts. This is Tomb 36, which is in grid
XVIII C, the only tomb in that square. I checked all the objects published
over the years that were identified either as from grid XVIII C or Tomb 36, in
many cases matching them with the 1983 list, and present the results here
with Negahbans publication date references. Tomb 36 contained at least:
four gold vessels (1964, figs. 111, 112, pl. VIII A; 1983, G6, 10, 11, 12); one gold
male bust (1964, fig. 101, pl. X); one gold bracelet (1964, fig. 77); a gold needle
(1964, fig. 42); a gold ring (1964, fig. 81); a gold seal (1977, 94, fig. 9); a gold
chain (1964, fig. 127); two bronze female figurines (1964, fig. 99; 1979,170171,
pls. 33: 2, 3; 34: 1, 2); three bronze seals (1977, 99, 100102, figs. 19, 20, 25, 26);
a bronze vessel (1964, fig. 28; 1983, B28); a bronze cymbal (1964, fig. 54); a
11 This conflation of grid and actual tomb number has caused confusion among scholars:
e.g., Peter Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Walter de Gruyter
& Co.: Berlin 1969) 56, 101, 104 cites Grab XIII B, Grab XXIIIc [misprint for XXIII G],
inadvertently mixing grid numbers with tombs.
Figure 2. Fibula from Marlik Tepe, Tomb 36.
bronze tweezer (1964, fig. 37); a bronze stag figurine (1964, fig. 96 A); seven
male and female terracotta figurines (1964, pl. XI; 1979, 159161, pls. 2931); a
terra-cotta charioteer (1964, fig. 94); a number of terra-cotta animal figurines
(1964, fig. 93, pl. IX).
Found with these objects in Tomb 36 was a Near Eastern bronze elbow
fibula, mentioned by Negahban in the aforementioned list of 1983 (page 90),
and which through his exemplary generosity and concern for scholarship
(and being aware of the intent of this paper) is published here for the first
time (fig. 2). The fibula has a wide triangular bow with a flattened bead
flanked by two discs or ribs, and a catch in the form of a human hand with
the thumb separate. It easily fits within Stronachs Near Eastern Type III, 7
group, where many parallels are forthcoming both in form and hand catch
with an isolated thumb. Close and relevant parallels are to be found in
Syria and Assyria, and, significantly, at sites in Iran.12 All, without exception,
12 Type III, 7, has, as Stronach pointed out, op. cit. (in note 2) 197, many different varieties,
and there are many comparisons readily made with the Marlik fibula. I call attention here
only to a few, which are closest in form: Deve Hyk, Moorey, op. cit. (in note 4) fig. 14:350,
dated to the 8th or to the 5th century bc; Nimrud, Stronach, op. cit. (in note 2) pls. L, 7, 10
(note the catch), LI, 13, and 912 for the catch; Gerar, Stronach, ibid. fig. 9:5, 6; from Iran
(excavated) see Godin Tepe, Bisitun, Susa, and sites in Luristan, vanden Berghe, op. cit. (in
note 6) fig. 11:5, 6, 7, 24; and fig. 9:9, 12, 15, 17, 18, 22 for the catch.
are dated in the late 8th or 7th centuries bc, or even later. Apparently no
other tomb at Marlik contained a fibula, but I maintain that because of the
presence of one in Tomb 36 there can be no doubt that at least this tomb was
closed not earlier than the late 8th century, and possibly in the 7th century,
i.e., in the Iron Age III period of NW Iranian chronology.
In all the publications concerning Marlik, Negahban has maintained that
the chronology of the cemetery is to be perceived as beginning sometime in
the 14th or 13th centuries bc and terminating in the 10th century, presum-
ably early in that century.13 In this posited chronology he has been followed
by a large number of scholars (too many to cite here), who over the years
have accordingly dated either the excavated materials from Marlik or many
related, but plundered or recently forged, objects from the antiquities mar-
ket to the late 2nd or, more rarely, to the early 1st millennium bc.14 It is not my
intention, considering the limited aims of this paper, to comment on the full
chronological range of all the Marlik tombs, except to note in passing that
some contained unbridged spouted vessels of Iron Age I, 2nd-millennium
13 Op. cit. (notes 9 and 10 for references): 1964, 3738; 1965, 312, 326; 1979, 157, 173; 1981, 369;
1983, viii, 95. In 1965, 326 Negahban posited that the Assyrians about the ninth century bc
or the Urartians a little later, brought Marlik to an end. In 1983, viii, 95, it is the Assyrians
who do this in the 10th century bc. There was, however, no Urartian penetration, neither as
settlers nor as raiders, into western Iran beyond Azerbaijan, many miles from the Caspian
Sea, as W. Kleiss has painstakingly demonstrated: see for example his Topographische Karte
von Urartu (AMI Ergnzungsband 3: Berlin 1976). Nor did the Assyrians at any time extend
their control, let alone penetration, so far east in Iran. What is more, the late date of Tomb 36
makes such a high date for the end of Marlik totally untenable.
Note, in the context of Urartian-Caspian relations, that Guitty Azarpay, Urartian Art and
Artifacts (The University of California Press: 1968) 12, 16, 82, note 28, pl. 1, refers to two Urartian
horse blinkers in the Feroughi collection, which she casually informs us derived from a
recent discovery in northwestern Iran , but which in fact derive from uncontrolled,
unwitnessed plunderers. This uncritical, and incorrect, attribution follows from a fictional
creation of Roman Ghirshman, Deux Oeillres en bronze des Rois dUrartu, ArAsiae 27:
1/2 (19641965) 5758, that the blinkers, inter alia, were discovered dans une des tombes
des montagnards de la rgion du sudouest de la mer Caspienne, and that the discovery
vouchsafes hitherto unknown information on Urartian-Caspian relations. The objects, of
course, do no such thing: they have no provenience, nor do we know what was found together
in situ or assembled in a dealers shop; see G. Gropp, Ein Pferdegeschirr und Streitwagen aus
Urartu, IranAntiq 16 (1981) 95, note 1.
14 The only exception of interest to this acceptance of chronology is Ekrem Akurgal,
Urartische und Altiranische Kunstzentren (Ankara: 1968) 67, 103106, who dates all the Marlik
material and that from Hasanlu and elsewhere in Iran to the 7th century bc, on the basis of
a preconceived notion that most of Iranian art is dependent on that of Urartu. Paradoxically,
Negahban cites modern forgeries in his lists of alleged parallels for his excavated pieces: 1983,
14, note 33; 19, note 71; 29, note 141; 58, note 12; 61, note 29.
form, which indicates that these tombs may date to that period.15 Nor is it
the time and place to initiate discussion with regard to the absolute date
of manufacture of all the objects deposited in Tomb 36, some of which
could conceivably have been curated heirlooms. Creative and intelligent
commentary cannot occur until all the tombs are fully published in discrete
groupsand even then all these problems will not be resolved. Neverthe-
less, the recognition that an Iron Age III, 8th7th-century bc tomb exists
in the Marlik cemetery, an occurrence hitherto unperceived, will inevitably
profoundly alter our conception of the chronology of a number of Marlik
and Iranian artifacts of similar or related styles, primarily vessels and fig-
urines. And because of the existence of Tomb 36, archaeologists will have to
consider the cultural and political implications of an unplundered (in antiq-
uity) cemetery, in which many rich tombs were built for a time spanning
apparently four or five hundred years in a peaceful and protected environ-
ment.
One example of a downward shift in chronology that results from the
lowering of the Marlik terminal date is of special interest and warrants
brief review here. In 1972 I published a superb bronze vase that had been
purchased on the antiquities market, and which I dated to a time between
1000800bc; my conclusions were in part based on the traditional Marlik
chronology.16 The vessel has six horizontal friezes or rows of animals that, as
I stated, strikingly parallel in form and style those represented on some of
the Marlik vessels. Since 1972 two scholars have suggested alternative dates
for the vase. Calmeyer17 believes that it was made in the 2nd millennium
bc, whereas Moorey18 suggested (but without discussion) that it should be
placed in the Iron Age III b period; in a subsequent publication he dated it
between 1000 and 650 bc.19 Four separate dates have thus been suggested for
the time within which the vase was manufactured, a clear indication that
15 Tombs 27, 45, 50, 52; Negahban, 1983 op. cit. (in note 10) S21, B4347. There is as yet no
evidence that the unbridged spouted vessel, the classic Iron Age I form, continued into the
Iron Age II period.
16 Oscar White Muscarella, A Bronze Vase from Iran and its Greek Connections, Met-
MusJour 5 (1972) 2550, figs. 111. Note that on pages 43 and 44 I expressed even at that time
some feeling that a 2nd-millennium date for all the Marlik material was not necessarily cor-
rect; Negahban (1983 op. cit. in note 10) neatly confirms that anticipation.
17 Op. cit. (in note 11) 205, note 442, c.
18 P.R.S. Moorey, Some Elaborately Decorated Bronze Quiver Plaques made in Luristan,
c. 750650 B.C., Iran 13 (1975) 25.

19 In Glenn Markoe, ed., Ancient Bronzes, Ceramics, and Seals (Los Angeles County Art
Museum: 1981) 84.

we are less secure about the chronology, and the archaeology, of NW Iranian
material than is usually recognized.
Now, Tomb 36, the fibula tomb, contained two gold vessels20 decorated
with two rows of animal friezes, and they were mentioned along with oth-
ers in my 1972 list of comparanda for the bronze vase. Whatever the date
one may assume for the manufacture of the vessels, they were in use and
culturally available up to the time of the tombs closure. This pertinent infor-
mation, this lowering of the lower limit for the presence of the Marlik ani-
mal frieze on vessels, signifies that the Metropolitan Museums bronze vase
could have been made at a time later than the early 1st millennium bc. This
date will have been, I suggest, sometime between, to give the upper limits,
the 10-9th century and the years close to ca. 700bc.21 In the original publica-
tion of the bronze vase22 I argued that the form of horizontal friezes that it
shares with the Marlik vessels may have been the ultimate prototype for the
origin of certain East Greek animal friezes represented on pottery vessels,
those of the Wild Goat style. The problem outstanding at that time with
this argument was that the Greek examples did not appear before the mid-
7th century bc, at least a century and a half later than the lowest date I had
then assigned to the vase. But now, given the extended chronology at Mar-
lik that manifestly indicates a longer life for the animal frieze in Iran than
hitherto recognized, the chronological gap between the advent of the Wild
Goat style and the Iranian parallels is significantly narrowed, if not indeed
closed. A single fibula has taken us on a long and promising journey, one not
yet ended.
We now turn to the second publication to be reviewed, this one concern-
ing the contents and date of closure of but one of the hundreds of tombs
excavated many decades ago at Assur in Assyria. In this instance a fibula
recovered in a tomb was not so much ignored, as was the case at Mar-
lik; rather it has been unappreciated for the correct chronological message
it bears. And if the aberrant chronological conclusion presented there is
20 Negahban, 1964 op. cit. (in note 9) figs. 111, 140, 144, pls. VIII A; XII; idem, 1983 op. cit. (in
note 10) GI0, 12; see also the Tomb 36 listing, supra.
21 The upper limits for the time of manufacture still obtain because of the presumably
earlier dates of the other Marlik vessels with animal friezes. Also relevant is the existence at
9th-century bc Hasanlu of a small group of ivory plaques that depict winged bulls executed
in the same style and body decoration as found on the bronze vase, Oscar White Muscarella,
The Catalogue of Ivories from Hasanlu, Iran (University Museum: Philadelphia 1980) 108115,
190191, numbers 214220. I still think it probable that the Metropolitan Museums vase was
made before 700bc.
22 Op. cit. (in note 16) 4349.
allowed to remain unchallenged, an archaeological distortion will result and

speculation will undermine controlled methodology.
In 198223 I argued that Gruft 30 at Assur24 could not be dated to the
9th century bc as assumed by R.S. Young and others,25 a conclusion that
had been held because of the presence of an inscription on a bronze bowl
mentioning one Assurtaklak, the same name borne by the Assyrian eponym
for 805 bc. Crucial to the argument was the presence in the tomb of a bronze
elbow fibula and a bronze lamp, both artifact types known from excavations
in Assyria and Iran to date to the late 8th century bc and later. These
arguments continue in force, all the more so in light of the evidence brought
forth in the recent article on the tomb by D. Surenhagen.26
After a detailed analysis and reinterpretation of the stratigraphy of the
tomb, considered to be under three distinct levels of private houses, Sren-
hagen concludes that it must have been constructed between 1075 and
824 bc. He further concludes that the overlying houses are earlier than the
excavators suggested date of the 7th century bc and could conceivably have
been built in the 9th century bc; the Gruftbelegung will have occurred before
this time.27 Following upon this is a chronological investigation of the tombs
contents, one claimed to be independent of the archaeological and strati-
graphical reconstruction. I do not think it an injustice, however, to state that
Srenhagens reconstruction and reinterpretation of the time frame for the
construction and closing of the tomb in fact did determine his conception of
the chronology of the Beigaben, for throughout his investigations he seems
to have held in mind the putative 9th-century closing, and all the tombs con-
tents are therefore furnished with an earlier life than they had in historical
reality.
Gruft 30 contained four skeletons; three earlier burials were placed to-
gether in the NE comer, presumably at the time of the manifestly latest
burial, which occupies the tombs center. The bronze lamp, which is not
discussed by Srenhagen, was found with the earlier burials. The fibula,
however, belongs with the latest burial.28 The Assurtaklak bowl published
by A. Haller is now clearly recognized as also belonging to the latest burial,
23 Muscarella, 1982 op. cit. (in note 5) 8.

24 Haller, op. cit. (in note 4) 109110, pl. 22.
25 Young, op. cit. (in note 5) 198, 234, 270.
26 Dietrich Srenhagen and Johannes Renger, Datierungsprobleme der Gruft 30 (Ass. I i
190) in Assur, MDOG 114 (1982) 103128.

27 Ibid. 110115, 128.
28 Ibid. Abb. 11, numbers 6, 35.
as Surenhagen neatly demonstrates by examining the original drawing of

C. Preusser, which differs from the one published by Haller that implies
it was with the earlier burials.29 Furthermore, Srenhagen has discovered
that the bowl held to the chest of the skeleton of the last interment is
also inscribed with the name Assurtaklak.30 Thus, two bowls inscribed for
Assurtaklak were deposited with the fibula in the latest burial.
In an important contribution for historical and chronological problems,
Srenhagen and the philologist J. Regner forcibly demonstrate that the
Assurtaklak of the bowls cannot possibly be the same man who was eponym
in 805 bc, as had been assumed for years, a conclusion I anticipated in 1982.
They significantly add Hiermit entfllt das Hauptargument fur die bish-
erige Datierung [i.e., 805 bc].31 Indeed, this states the case succinctly, and
it allows archaeological research to proceed without preconceived notions
or restraints. But one restraint is replaced by another. For, inasmuch as the
reinterpreted 9th-century bc closing date for the tomb is presented as a real-
ity, Srenhagen is consequently able to conclude that all the Beigaben must
have been deposited in the 9th century, and probably within the first half of
that century.32
What of the fibula and how is its presence reconciled with a 9th-century
closure date? Srenhagen, like Stronach,33 identifies it as Type III, 6, a type
that, as reported above, had its advent in the Near East not prior to the late
8th century bc. But in his attempt to avoid the chronological implications of
the fibula, Srenhagen rises above the level of the evidence and compares it
in a facile manner with Stronachs East Mediterranean Type II group, those
with arched or round bows.34 We need not become involved in a discussion
of the chronology of this group except to state that it, too, fails to support
a 9th-century bc date for the Assur fibula, and it is not relevant: it has a
29 Ibid. 116, Abb. 11, 124; Haller, op. cit. (in note 4) plate 22, c-e.
30 Srenhagen and Renger, 103108, Abb. 15.
31 Ibid. 128.
32 Ibid.
33 Stronach, op. cit. (in note 2) 196197, fig. 8:2; actually, the fibula could be considered as
fitting into Type III, 7, as the central bead seems to be flanked by two smaller ones: but this
is not relevant to chronology.
34 Loc. cit. (in note 26). Ich vermag nun keinen gravierenden Unterschied zwischen
dieser Fibelgruppe [Type II, 4] und D. Stronachs Gruppe III.6 festzustellen, zu der auch
die Fibel II 190 al [from Gruft 30] gezhlt wird, und schliesse ein lteres Entstehungsdatum
als das dart angegebene (7. Jhdt. v. Chr.) nicht aus. In this context Srenhagen refers to the II,
4 fibulae at Megiddo (ca. 1000800 v. Chr.), for which see here note 2. He also accepts Hallers
clearly tentative division date of 824 bc for the alleged time when round bowed fibulae were
replaced by triangular ones, for which see here note 4.
different form of bow than Type III, 6. Srenhagen then goes on to cite only a
single excavated parallel of a triangular fibula. The sole parallel offered is not
without relevant interest in a chronological discussion, and its introduction
reveals far more than was intended. It is a Type III, 7, example from War
Kabud in Luristan, Iran. Without discussion and presentation of evidence
Srenhagen suggests this site need not be 8th7th century bc in date, but
could be 9th century. In fact, War Kabud, on the basis of all the information
available to Iranian archaeologists, must be an Iron Age III site, i.e., 8th7th
century bc in date;35 and of all the other pertinent excavated parallels for the
Assur fibula, not a single one predates the late 8th century bc.36
Turning now briefly to the bronze lamp found with the earlier burials in
Gruft 30, a fine parallel was excavated at Baba Jan in western Iran from an
intrusive horse burial. Its date of deposition cannot be earlier than the late
8th century and is most probably in the 7th.37 Here, then, we have another
important piece of evidence for a post-9th-century bc date for the Assur
tomb, one not considered by Srenhagen. And Srenhagens chronological
discussions for other objects in the tomb, e.g., comparing a knobbed bronze
bowl to a different, lobed, example from War Kabud, or dating neo-Assyrian
cylinder seals only to their upper limits, do not in any objective sense
indicate a 9th-century date for the tomb or the fibula.38
What of Srenhagens conclusion that the tomb was closed in the 9th
century bc and how secure is it? I am forced to suggest from an examina-
tion of the finds that his reconstruction of the stratigraphy is not accept-
able and warrants review. I also believe that it will develop that Es besteht
somit kein Widerspruch zwischen baugeschichtlichem und archologisch-
ern Befund.39
Gruft 30 and its contents have not altered our opinion based on empirical
documentation that the Near Eastern fibula is a Leitfossil, an object of fun-
damental importance for dating material juxtaposed in tombs and secure
stratigraphical levels to the late 8th century bc or later.
35 Ibid. 124125, note 63. See vanden Berghe, op. cit. (in note 6) for the chronology of War
Kabud and other Luristan sites; this important paper is not cited by Srenhagen.
36 For references see note 12, and especially Moorey, 1980 op. cit. (in note 4), fig. 14, no. 350.
37 Clare Goff Meade, Excavations at Baba Jan 1967: Second Preliminary Report, Iran 7
(1969) 123126, fig. 7; on page 126, notes 29, 30, the Gruft 30 lamp and fibula are cited. I plan
to write elsewhere on the chronology of Baba Jan but will note here that I firmly hold the
view that the site did not come into existence before the 8th century bc (Period III), and that
Period II is not earlier than the 7th century bc. If correct, this view supports the dating of the
horse burial to the 7th century bc.
38 Op. cit. (in note 26) 125126, 128.
39 Ibid. 128.
chapter twenty-eight
PARASOLS IN THE ANCIENT NEAR EAST*
The earliest surviving examples of sunshades or parasols represented in

ancient art appear in Fifth Dynasty Egypt (c. 2450 bc), when they are first
associated with males, and subsequently (Dynasty XI and later) with both
males and females. Earlier representations of fan-shaped forms attached
to shafts are assumed by some scholars to have been sunshades, but this
is not certainthey could have been fans.1 Sunshades began to be rep-
resented a little later in the ancient Near East, where they were always
portrayed in association with a (male) king. In Egypt and the Near East,
the monarch himself never held the shade; rather, an attendant positioned
behind him held the shaft so that the shade was placed above the kings
head.
The earliest preserved Near Eastern example appears on a fragmented
Akkadian stele depicting Sargon I, king of Akkad (c. 2310bc), in a victory
celebration (Fig. 1).2 Behind the king, an attendant holds up a parasol with
a curved shade that is placed close to, but not directly over, the kings
head (the rigid registers probably prevented an accurate depiction of this
gesture). Two oblique struts flank the centrally placed shaft, but whether
this indicates that the shade could be collapsed is not known. The shade
is curved, a form that eventually became canonical in the Near East; it
differs from the traditional rectangular, flat shade documented in Egypt
from Dynasty V.
The next extant representation of a sunshade occurs on an unexcavated,
unprovenienced cylinder seal (Foroughi Collection); it is probably Assyrian,
thirteenth century bc. Depicted is an attendant holding a flat-shade parasol
* This article originally appeared as Parasols in the Ancient Near East, Source 18, no. 1
(1999): 17.
1 See, for a review and illustrations, H.G. Fischer, Sunshades of the Marketplace, Jour-
nal of the Metropolitan Museum of Art 6 (1972):151156; id., Sonnenschirm, in Lexikon der
gyptologie 5 (Wiesbaden: 1984).
2 The stele was recovered in Susa, Elam, whither it had been brought as booty in the
twelfth century bc; Anton Moortgat, The Art of Ancient Mesopotamia (London: Phaidon,
1969), pl. 125.
818 chapter twenty-eight
Fig. 1. Stele of Sargon I of Akkad (c. 2310bc). Susa. Louvre, Paris.
above the head of a seated king (Fig. 2).3 A body-length shaft shown with
upper struts supports the shade from one side, not from the center, and a
long, narrow extension curves down from the rear of the flat shade, enclos-
ing the attendant.4 Perhaps this flat shade reflects the Egyptian form. The
latter, however, has its shaft centered, not placed to one side, and there is no
long rear element.
3 Foroughi Collection. E. Porada, Aspects of Elamite Art and Archaeology, Expedition 13,
nos. 34 (1971):33, fig. 9 (considered by her to be Elamitebecause it is owned by a modern

Iranian?); Donald M. Matthews, Principles of Composition in Near Eastern Glyptic of the Later
Second Millennium B.C. (Gttingen: 1990), pp. 111 f., no. 514, considers it to be Assyrian.
4 We might call this form a Sonnendach, a term used by E. Unger in his (still valuable) arti-
cle Sonnenschirm, in Reallexikon der Vorgeschichte 12 (Berlin: 1928), pp. 308310, esp. 309;
to describe an Assyrian curved form of sunshade, see below. The curved unit seems to form
a canopy-baldachin over the kings space.
parasols in the ancient near east 819
Fig. 2. Cylinder seal. c. thirteenth century bc. Collection Foroughi, Porada

Archives, The Pierpont Morgan Library, New York. (Photo: Edith Porada.)
Parenthetically, pottery representations show that in the fourteenth cen-

tury bc the parasol with a curved shade was represented in the Bronze Age
Greek sphere, here, too, serving as a status symbol reserved for ranked males,
possibly kings. This ensemble of parasol and ideological message originated
in the Near East or Egypt given the shade form, probably the former. After
the early twelfth century bc, there is a gap in the representation of parasols
in the Greek sphere until the sixth century bc, at which time they are given
a new and different meaning (below).5
Parasols next appear (better, are again extant) centuries later in late
Assyrian art, from the early ninth through the seventh centuries bc (Fig. 3).
They are very commonly represented during this period. Beginning with
Ashurnasirpal II (883859bc), the kings parasol, with a curved shade
5 See M. Miller, The Parasol: An Oriental Status-Symbol in Late Archaic and Classical
Athens, Journal of Hellenic Studies 112 (1992):95 and pl. 1a, and bibliography in nn. 811.
Fig. 3. Relief of Ashurnasirpal II (883859bc).

Nimrud. British Museum, London.
precisely like that of Sargon I depicted a millennium and a half earlier

(Fig. 1), is held above his head when he is engaged in parade, in battle, in
sacrifice, or performing palace duties. It was also carried in or alongside his
chariot, sometimes attached to a fixed pole; it is depicted on wall reliefs,
bronze gates, ivory carvings, obelisks,6 and on seals illustrating ritual scenes.7
Beginning with the earliest representations and continuing to the elaborate
examples with tassels on the seventh-century-bc reliefs of Sennacherib and
Esarhaddon, Assyrian parasols have upper multistruts connected to a knob,
surely indicating here that the parasol was collapsible.
Also depicted on the ninth-century-bc Balawat Gates of Shalmaneser III
is a form rendered as a shade curving out as an extension of the shaft. If this
form had a special charged value (a type of baldachin?), we do not know,
6 For example, R.D. Barnett, Assyrian Sculpture in the British Museum (London: 1975), pls.
27, 35, 37, 45, 65; id., Assyrian Palace Reliefs (London: Batchworth Press Ltd., n.d.), pls. 141, 158,
159, 163; J. Reade, Assyrian Sculptures (Cambridge: Harvard University Press, 1983), figs. 44, 62,
98; J. Br, Der assyrische Tribut und seine Darstellung (Neukirchen Verlag, 1996), p. 215, n. 1677,
Abb. 53, 54.
P. Calmeyer in Altiranische Bronzen der Sammlung Brckelschen (Berlin: 1964), pp. 50f.,
Abb. 8, published as ancient Assyrian an unexcavated bronze vessel decorated with a frieze
of motley, allegedly Assyrian troops attacking a fortress. One of the figures casually car-
ries a parasol over his shoulder, another at an angle behind a human-yolked vehicle; no
king is manifestly represented. Such depictions are impossible compositions in known
excavated(!)Assyrian iconography. I suggest that the scene is surely a modern addition to
an ancient vessel, a modern forgery, not an ancient Assyrian scene. The corrosion should be
tested.
7 S. Herbordt, Neuassyrische Glyptik des 8.-7. Jh. v Chr., The Neo-Assyrian Text Corpus
Project (Helsinki: University of Helsinki, 1992), Taf. 20, nos. 2, 3.

but it seems to be a Sonnendach held above the king in the same manner as
a parasol.8
Assyrianizing scenes decorating late eighth-seventh-century-bc Phoeni-
cian metal vessels depict a king with a parasol, emulating Assyrian royal
practice.9 Absorption of canonical Assyrian iconography and ideology is
also evident on mid-seventh-century-bc Urartian royal seals depicting King
Rusa II attended by a parasol bearer.10 These Urartian parasols were either
plain or had tassels (or bells?) at the rim (like those from contemporary
Assyria). While no earlier Urartian representations are extant, it is proba-
ble that royal-parasol ideology was practiced there earlier.
A little-known representation of a parasol exists in an eighth-century-bc
North Syrian context; it is the only known example in the art of that area.11
A stone funerary stele from Marash depicts a facing seated couple. Rather
than held by an attendant, the parasol extending over the male figures head
is shown as if its shaft is a curved extension of the chairs backa unique
depiction. This Assyrianizing iconography, locally modified, surely informs
viewers that the male is a king (or member of the royal family?).
A wooden knob with the struts of a collapsible parasol was excavated
at Gordion, the Phrygian capital in central Anatolia,12 from a childs tomb
(Tumulus P), dated to the late eighth century bc. We do not know whether
this example was locally made or came to Gordion as a gift from an Assyrian
king (Sargon II). Nor do we know the extent to which Phrygians employed
parasols since no Phrygian representations of them exist. Given, however,
the gender-determined function of the parasol, Tumulus P surely was the
burial place of a prince (or a prematurely deceased king?).
The sliding knob section with eight rectangular cuttings belonging to
a wooden parasol was excavated at the Hera temple on Samos, where it
had been dedicated (probably by a non-Greek) sometime in the seventh
century bc.13
8 Barnett, Assyrian Palace Reliefs, pls. 158, 162; Unger, p. 309. Note that the typical parasol
form is also depicted on these same reliefs.

9 G. Markoe, Phoenician Bronze and Silver Bowls from Cyprus and the Mediterranean
(Berkeley: University of California Press, 1995), nos. E2, E11, E13.

10 U. Seidl, in W. Kleiss et al., Bastam, II (Berlin: 1988), p. 146, pls. 22, 23.
11 It was only recently published by A. Muhibbe Darga, Ge-Hitit Dnemi Maras Mezar
Stellerinden mek ve Gzlemler, in Light on Top of the Black Hill, Studies Presented to
Halet ambel (Istanbul: 1998), pp. 243, 249, figs. 2b, 8.
12 R.S. Young et al., Three Great Early Tumuli (Philadelphia: University of Pennsylvania,
University Museum, 1981), Fig. 45, Pl. 32 F, G.

13 See H. Kyrieleis, The Relations between Samos and the Eastern Mediterranean: Some
The evidence manifests that from its incipience, the ancient Near Eastern
parasol was gender- and socially oriented. The parasol and its bearer were
restricted to the kings person, his royal prerogative. It was not in common
use socially; it was not used in religious ceremonies; it functioned solely as
a marker of the kings rule and authority.14
This prerogative continued without change into the Achaemenian period
(late sixth-fourth centuries bc), and at Persepolis several kings, accom-
panied by a parasol attendant, are depicted in doorways (that is, mobile;
Fig. 4).15 When the king was portrayed enthroned (that is, static), however,
a baldachin was fixed in place and covered the full throne area.16 Probably
the baldachin, which had already been represented in some royal scenes
in Assyrian times (i.e., on the Balawat Gates),17 became associated with the
charged meaning of the parasol, developing into its fixed counterpart, one
that shaded, protected, and defined the space around the king, not merely
his body. Parts of an iron parasol are reported to have been excavated at
Persepolis but remain unpublished.18
In Lycia, at Trysa and Xanthos, parasols were depicted on grave reliefs;19
and at Kizilbel, a tomb painting depicts a parasol bearer on a ship, this
parasol held by an attendant facing the protected figure.20 In this western
region of the Achaemenian Empire, a local, not canonical, Achaemenian
concept obtained, for the representations are manifestly not those of kings,
but of high-ranked males, possibly from satrap families,21 and a funeral role is
Aspects, in The Civilizations of the Aegean and Their Diffusion in Cyprus and the Eastern
Mediterranean, 2000600 B.C., ed. V. Karageorghis (Larnaca: 1991), p. 131, pl. XXX: 2; and
Kyrieleiss forthcoming article in Anadolu 23. See also Miller, 95, n. 24.
14 See U. Magen, Assyrische Knigsdarstellungen Aspekte der Herrschaft, Baghdader For-
schungen 9 (Mainz am Rhein: 1986), p. 39 and chart on p. 131; Br, p. 215.

15 E.F. Schmidt, Persepolis, I (Chicago: 1953), pls. 7576, 138139, 178181. I could find
no parasol scenes represented on seals. Note that the use of a parasol by Persian kings is
another example of iconographic and ideological continuity from Mesopotamian culture of
the previous era.
16 Ibid., pls. 7778, 104107.
17 Barnett, Assyrian Palace Reliefs, pl. 142; see also n. 8, above.
18 Personal communication from Judith Lerner, from a third source. It is also mentioned
by R. Ghirshman in LArt animalier aulique achmnide, Monuments et Memoires 60

(1976):12, n. 3.
19 B. Jacobs, Griechische und persische Elements in der Grabkunst Lykiens zur Zeit der
Achmenidenherrschaft (Paul strm Frlag, 1987), Tafs. 6, 15.

20 M. Mellink, Excavations at Karatas-Semayk and Elmali, Lycia, American Journal of
Archaeology 75 (1971):248 f., pl. 52.

21 Also noted by Miller, 94 f. See also the sixth century bc so-called Satrap Sarcophagus
from Cyprus, J.L. Myres, Handbook of the Cesnola Collection of Antiquities from Cyprus (New
York: 1914), pp. 228, 230, pl. 1365A, which seems Assyrianizing.
Fig. 4. Relief of Persian king Darius I (522486 bc). Persepolis.

(Photo: Lewis Callaghan; gift of Mrs. Lewis Callaghan, 1989.)
documented. Incidentally, this is another example of Achaemenian political

tolerance of local customs.
Perhaps the chronological gaps in the representations known to us that are

documented above are due to archaeological accident, not to discontinu-
ance and reintroduction of royal iconography. That is, the ideology inform-
ing the parasol function probably existed in areas of the ancient Near East
continuously, certainly from Akkadian times through the Achaemenian.
(But what of the Hittites, Hurrians, and others?) The royal function of the
parasol may have developed from an original intent to protect the king from
the sun that grew into a sense both to define and protect his unique space.
As noted, the parasol, and its association as a ranked, male attribute, was
assimilated by the Greek sphere in the late Bronze Age. After a hiatus, it
was again depicted in the classical age, probably during the Achaemenian
period (the Samos example is not evidence for western Greek use at that
time). These later (new orientalizing) depictions, however, reflect a casual
use, with no evidence of Near Eastern ranked function. Furthermore, the
parasol was now commonly used by women. It was also shown carried by
effeminate men, a motif to evoke ridicule of their behavior, considered to be
normal of Persian men, not Greeks.22 The parasol had thus become a more
democratic fashionable accessory, but it continued to be a visual marker to
delineate social attitudes within society.
22 For a good summary of the classical evidence of parasol function, see Miller (n. 5,
above).
Section Three
The Antiquities Market and the Plunder Culture

chapter twenty-nine
THE POPE AND THE BITTER FANATIC*
Abstract: The paper contrasts the professional activities of Professor Ezat Negah-
ban with that of Arthur Upham Pope. Beginning as a young man and continuing
throughout his career as a field archaeologist, Negahban combated the plundering
of archaeological sites in Iran, which he knew resulted in the destruction of ancient
Iranianand the worldscultural history. In this endeavor, he was opposed by
many forces, dealers, collectors, members of the royal family, and even alleged schol-
ars, among them Roman Ghirshman. But he persevered.
One of Negahbans opponents was Pope, an extraordinary, and complex individual
who played a large role in bringing forth information about ancient and Islamic Iran
through his involvement in exhibitions, institutions, and publishing. Paralleling
these activities, however, was that of Pope the dealer in Iranian antiquities, a role
which he deliberately kept secret from scholars. It is argued that selling antiquities
was the main force in his life, that it informed and manipulated his other, better
known, activities. The paper concludes that the beliefs and activities of the two
men represent the two opposing forces confronting archaeology today, excavation
or destruction.
In addressing Negahban you walked right into the tigers den. He is a bitter
fanatic playing the artificial role of protecting, for the Persian people, their
rightful heritage of Iranian antiquities of which they are being shamefully
robbed by the antiquarians (all criminal thieves) and their criminal accom-
plices, the greedy and unscrupulous museum directors. He had prepared a
rather violent resolution to be passed, he hoped, by the congress Ghirsh-
man and I voted against it and 300 sheep voted for it.
This animadversion was written by Arthur Upham Pope to one of his dis-
ciples, Rexford Stead (Gluck and Siver 1996: 537). The congress mentioned
was the Vth International Congress of Iranian Art and Archaeology meeting
in Iran in 1968, the year the letter was written, and the violent resolution
was in fact a moderate appeal: given the growth of clandestine and ille-
gal traffic in antiquities which destroys both the cultural and historical
* This chapter originally appeared as The Pope and the Bitter Fanatic, in The Iranian
World: Essays on Iranian Art and Archaeology Presented to Ezat O. Negahban, eds. A. Alizadeh,
Y. Majidzadeh, and S. Shahmirzadi (Teheran: Iran University Press, 1999), 512.
828 chapter twenty-nine
heritage of mankind, the congress appeals to all countries and UNESCO

to devise and adopt effective ways and means to prevent the export from
and the import to one country of antiquities from another except through
official and authorized channels.
Popes sarcastic evaluation of Negahban and his disdain for the purpose
of the congress resolution, which Negahban played a large role in formulat-
ing,1 immediately set in my mind the theme for the Festschrift article I had
been seekingthe vital difference between the work and goals of Negahban
and those of Pope. One could not think of two more appropriate individuals
whose achievements better epitomize the two opposed systems that prevail
in the archaeological world at large. While one worked all his life to save
the past for the future, to preserve the planets history, the other incessantly
worked to destroy it, to reduce it to exhibitions of antiquities on museum
and collectors shelves, and to ridicule those who opposed these objectives.
The great virtue of Negahban was that he took his role as archaeologist seri-
ously, he respected archaeologys needs and protected the artifacts within
its care. Early in his career he recognized that he was engaged in a battle
with powerful forces, and it was a battle he has fought all his life, in the posi-
tions he has held, and in his writings and lectures; and Pope wasnt his only
enemy, others included archaeologists, such as Roman Ghirshman. Negah-
ban was one of the first archaeologists I had ever encountered who spoke
openly about the tragedy wrought by plunder and the role of museums and
collectors in this activity (not once in all my years as a student did anyone
ever mention, let alone explain, how objects reached museums and collec-
tions). And, to speak personally and appropriately, it was a lecture of his that
first turned me on to the problem (this and a lecture by Machteld Mellink
and the earlyand isolatedwritings of Clemency Coggins helped me to
understand the nature of collecting).
To this day many students and colleagues remain unaware of Negahbans
other important contribution to archaeology, that he was not only the exca-
vator of Marlik and other sites. With this in mind, I write this article as a
dedication to Negahban as an expression of my respect for his fighting the
1 In the Vth International Congress of Iranian Art & Archaeology, Teheran, 1968: xxxiv,
Negahban claimed that there were only two negative votes on the resolution; he refrained
from mentioning names. Pope says it was he and Ghirshman. This is contradicted by Gluck
who proudly informs us that he too, along with another individual, also voted against the
resolution (534, 540). What editor Gluck never informs the reader, thereby paying the best
compliment to his mentor by imitating him, is that he is, in addition to being a publisher, an
antiquities dealer operating from Japan.
the pope and the bitter fanatic 829
good fightand for being a decent man. I trust the offering will alert stu-
dents (I am not optimistic about many of their teachers) to the two opposing
forces exemplified by Negahban and Pope. Alas, they should realize that at
present the Negahbans are on the losing side.
Negahbans concern on these matters began to generate in the 1950s
before he excavated at Marlik (Negahban 1996: 5 f.), and accelerated there-
after, for in all his subsequent international lectures he never ceased to
contrast archaeologys work with that of its enemies. He went to Marlik in
the first place because he became aware that plundering was in progress
there, that plunderers, dealers, and smugglers were active. It turned out that
there was to be only one campaign at Marlik. For, after commencing exca-
vation he was forced to remain at the site for fourteen continuous months
(19611962) in order to save the site from destruction. This perseverance
was unprecedented, and I suggest it was heroic. From day one, he had to
confront intrigues by dealers, corrupt officials, and members of the royal
family, who collectively attempted to terminate his work so that they could
begin their work (which would not have taken fourteen months, merely a
few days). Only an appeal to the Shah allowed him to continue. Few schol-
ars are aware that he suffered a further, terrifying, ordeal, that his team was
physically attacked and his camp destroyed, and that soldiers were sent to
protect himfrom the Popes of the world. As he succinctly noted we
had to be continually on the alert to protect the excavation and even our
own lives (Negahban 1996: 10). Here we have a real life dramawith none
of the protagonists playing an artificial roleenacted by the two oppos-
ing forces; the performance is a paradigm of archaeology versus the plun-
derer/collectors. Negahban has written a professional and personal autobi-
ography, which, inshallah, we will soon read.
Arthur Upham Pope is a name known to all scholars of Iranian art. He
played a role in the organization of several international congresses on Ira-
nian art and archaeology, in Philadelphia (1926), London (1931), Leningrad
(1935), New York (1940), Iran (1968). He traveled to Iran many times, record-
ing countless records and photographs of Iranian monuments. He was the
founder (1928) of the Asia Institute (later adopting different names), which
had a faculty of several European refugees, and which sponsored archaeo-
logical excavations. He was an advisor (for purchasing antiquities) to sev-
eral museums. And, his best and most lasting work, he organized, edited,
and brought to fruition the first major publication of Iranian art ever pro-
duced in the west, The Survey of Persian Art (SPA 19381939) in six large
volumes. He was beloved by the Iranian royal family for his love of Iran
and for his contributions to disseminating knowledge of its rich heritage. In
1966, the royal family gave him and his wife a home in Shiraz, and after their
deaths (1969; 1977) a mausoleum and a park named for him in Esphahan.
This summary does not do justice to the activities and life of this truly
extraordinary, complex individual. To a large degree the full range of the
activities and life of Pope emerge in a biography published in 1996, Surveyors
of Persian Art, edited and compiled by Jay Gluck and Nol Siver. A large vol-
ume of 658 pages, it is a collection of, inter alia, selections from biographical
and anecdotal records, and selected letters, many written by or to Pope and
his wife, Phyllis Ackerman. Two prominent collaborators in the venture are
Robert Payne and Rexford Stead, both of whom were cooperating on writ-
ing a biography of Pope, but died before its completion; included are also
extracts from Popes unfinished autobiography which he called Nine Lines.
The collaborative biography is a labor of love, assembled and written
by Popes students and friends, all of whom loved, no, adored him (xix)
utterlya sentiment in evidence throughout. This love compelled the pub-
lication, it weighs heavily on all the collaborators and editors assessments
of Popes deeds and attainments, and determined many of their positive
evaluations that is often expressed against the evidence they themselves
present; many critical evaluations of Pope are perceived as an attack.
Nonetheless, there remains a positive, indeed, an honest, quality, which
enriches the value of the work considerably, inasmuch as the editors have
chosen (after much wrestling with themselves, i.e., 574 f.), to give us the warts
too. This openness now allows us to discern that, far more intensely and for
a much longer time than hitherto realized, Pope was involved in the antiq-
uities market. He was in fact an active and aggressive antiquities dealer.
And the published evidence is there to demonstrate that this role was the
core of his life, was the motivating factor that informed most of his other,
apparently scholarly, activities. To appreciate the full thrust of this charge,
one must read all the published evidence; here I present appropriate high-
lights.
Like Negahban, Pope began his life-long career early. The earliest evi-
dence of Pope as a dealer seems to be 1908, when he began to sell rugs in
Providence, Rhode Island (49); he still listed himself as a rug dealer as late
as 1921 (as seen on a letterhead: personal knowledge). That he (along with
his wife) was dealing in antiquities at least from the mid-1920s is manifested
from Popes correspondence (154ff., in a chapter labeled Some Correspon-
dence on Dealingthe some indicating that only a sampling of the avail-
able information is presented). Included are letters written in the 1920s and
1930s about sales, prices, professional fees, etc., to the dealer Rabenou, with
whom he maintained a life-long business relationshipand whose family
called him Uncle Arthur: (155156, 170, 179f., 204, 207), and to many museum
directors, to whom he invariably bragged that his prices were considerably
cheaper than those of other dealers (adjusted to todays prices, his figures
would be multiplied considerably). Mentioned inter alia is his archaeolog-
ical collection (159written by Ackerman); that he has purchased over
three-quarters of a million dollars worth of Persian things (167); that I have
supplied more than $800,000 worth to more than twenty museums (413);
and a boast that I have contacts with eighty percent of the sources of Per-
sian art in Persia
In Nine Lives, Pope lists as one of his accomplishments that he was a
Purveyor of Persian art2 to several famous collectors and finally exclusively
to nearly a score of museums in both this country and abroad. Gluck (xv)
reports that from their Paris home Pope and Ackerman did art business
(notice the editorial euphemism slipped in here and elsewhere in the
biographyPope was in the art business, not the antiquities business); and
Stead reports (583) that one of Popes aides told him that Pope purchased
objects every trip he made to Iran. In 1968, just before he died, the Tehran
Journal reported that he was advising a young college student, one of my
[Pope] very newest students, how to purchase antiquities, and he called her
attention to one particular Tehran shop where he had purchased more than
a quarter million dollars worth of things (527).
While I had known for some time that Pope was a clandestine antiquities
dealer (Muscarella 1979: 5f., 13), the newly released information fleshes out
details of this activity significantly and blatantly.
Pope of course never revealed his role as a dealer in his publications and
within his institutional activities. But there was explicit public notice that he
and his wife Phyllis Ackerman owned antiquities, for they identified them-
selves (sometimes using one of their names, sometimes both) as owners of
a number they illustrated and recorded in allegedly disinterested scholarly
publications they controlled: SPA Vol. 1: Figs. 38, 65, 71; Vol. 7: Pls. 10, 33, 41, 59,
60, 62, 74, 111, 124, 192, 255, 256; in the 1931 London Catalogue of the Interna-
tional Exhibition of Persian Art: pp. 23, 8, 12, 1617; in Ackermans 1940 Guide
to the Exhibition of Persian Art: listing on page i, etc.; and in Popes Master-
pieces of Persian Art, London, 1945, Pls. 67, 27; and let it be noted that many
of the antiquities he published in all these publications had already been
sold by Pope (see also Illustrated London News (ILN) below). There is also at
2 Which title would have been most appropriate for the biography: did the editors not
realize that the title they chose would inevitably be smiled at?
least one published scholarly record that Pope sold antiquities to a museum,
here the University Museum, University of Pennsylvania (Legrain 1934: 8,
19); we can now cite the letters on this transaction published in the biog-
raphy (157f.). Worth relating in this context is that although Rexford Stead
acknowledges in a 1982 letter discussing my 1979 paper (169) that he knew
the SPA included objects in dealer possession and a great many others
owned by collectors who probably obtained them through Arthur when, in
the 1920s, he was actively involved in selling, he does not mention, what he
must have known, that one of the dealers was Pope.
Just as he manipulated readers in his major publications, Pope mani-
pulatedbrilliantly, one is inclined to observethe ILN. The editors and
the magazines readers, which included students and scholars, believed that
Pope was a disinterested scholar publishing recently discovered archaeo-
logical material (viz. ILN 5/6/39: 790, where the editor writes that Pope is
known for his discoveries of ancient Iranian bronzes). In fact, he was
publishing objects being offered for sale, or that he had sold, thus using
the ILN as a sales catalogue (see Popes ILN bibliography on pp. 613 f.; Mus-
carella 1979: 13). A clue to this activity exists in ILN 9/13/30:444, lower right,
where he specifically identifies objects illustrated as in the collection of
Mr. A. Upham Pope. A 1930 letter published in the biography (160) reveals
Pope bragging to the Boston Museum that a number of objects he sold them
will be illustrated in a forthcoming ILNjust part of his sales service. These
camouflaged sales catalogues were accorded authority by Popes presenting
himself as Director of the International Exhibition of Persian Art (viz. ILN
9/6/30: 292; 9/6/30: 388; 9/13/30: 444), and as Director of his Institute (ILN
5/31/41: 718; 3/1/41: 292).
Pope further used the magazine to enlighten us on archaeological activ-
ity in western Iran, to share his professional knowledge concerning what
archaeology and its techniques were. It is the equivalent of a semesters
education to compare Popes straight faced and casual description of how
Lurestan bronzes were discovered, how he reported what in reality was
the plundering and total destruction of the Lurestan culture that he was
financing, with Negahbans fourteen months work at Marlik and the subse-
quent activities he observed after he left the area. In the ILN of 10/22/32: 613
Pope introduces us to a Mr. Rahim, who with his brother was closely associ-
ated with the finding of the bronzes The brother spent ten consecutive
months, collecting bronzes Rahim also travelled to Piravand, where a
great quantity of fine and typical Lurestan metal was found, about a thou-
sand pieces in less than two months. Rahim was also present at the opening
of many graves When Nwgahban (1996: 11) returned to the Marlik area
after his excavations terminated, he counted more than 2000 ditches cut
by the plunderers; and that as many as 400 persons had been employed in
this clandestine digging, with the local gendarmes receiving a bribe of 5 rials
per worker to look the other way.
Pope used his close involvement in the international exhibitions he spon-
sored to sell his wares. This was already indicated by the illustration of
objects he owned in the publications of those exhibitions mentioned above.
More explicit information that these expositions were used by Pope as a
venue for his sales is now clearly divulged in the biography. In a letter to
Robert Payne in 1982 Rexford Stead wrote: Much of what was shown at the
Exposition [Philadelphia 1926] was actually for sale Much of the objects
on view had been lent by dealers, and even AUP and PA had works for sale
(577). In ano ther letter to Payne (578f.) written the same year, he states
that Pope arranged in 1926 on his own authority to organize a parallel expo-
sition at the Pennsylvania Museum (later the Philadelphia Museum of Art)
because Pope questioned whether the Exposition was the best vehicle for
such a presentation, whether it would attract the wealthy collectors AUP
wanted to reach.3
In a letter to a Museum Director (163) written in 1932, Pope offers for
sale an object that was one of the knockouts of the showthe show was
the 1931 London Exhibition. Pope also slipped in the sales pitch that the
Hermitage was negotiating for the piece, so this museum director had better
act fast to purchase. Pope himself, in his biography, refers to his exhibiting
in the 1931 Exhibition his own wares and those he had just sold (189).
As for Popes manipulation of his Institute to serve his antiquity dealing,
one need only refer to his use of his title Director of the Institute in his
ILN catalogues (above), and to what amounts to a confession by Stead from
a 1982 letter already cited above (169): Although his Institute association
doubtless [sic] enhanced his later [not earlier?] sales, Im confident that
3 In fact, there were individuals involved with the 1931 Exhibition, who charged that Pope
controlled the committee selecting objects for exhibition, that he was selling antiquities as
well as acting for dealers, and that he knowingly sold forgeries. A committee held hearings,
which disbanded when accusers refused to press their charges in public. Pope was exoner-
ated: see the chapter The Trial , pp. 205 ff. Payne (206) exclaims in his unfinished biography
Arthur was handsomely vindicated.
Although Pope was often accused of selling forgeries, I do not discuss that issue here
because I have no objective information that he did so knowingly. He certainly defended
himself openly and clandestinely against the charge (Muscarella 1979: 5f.). There is no doubt
that he sold forgeries, including the well-known Andarz Nameh manuscript.
he never used the nonprofit Institute illegally, yet he did personally get
involved, and this presents an ethical problem. No more need be said on
this issue.
Although the contributors to the biography often defend and rationalize
Pope the art dealer as one who only took commissions (see 19, 154 ff., 169,
206f., 216, 316, 574, 577), they deserve praise for sharing with us contrary
evidence as well as their misgivings about the man, some of which we have
already seen. Some years before 1981 Payne is quoted by a third party to have
said of Pope What a magnificent charlatan (408).4 Stead wrote to Payne
in 1979 is the picture [of Popes accomplishments] compromised by his
dealer aspect? This is troubling my conscience (574). Three years later he
wrote to Payne a cri de coeur (579): Robert, Im beginning to think that
all of us who were drawn to Arthur had some personality gaps or quirks
or weaknesses. Perhaps AUP sought out such people! We made him into a
father figure, possibly out of need. This says much about Pope, his devotees,
and the impulses behind the present biography.
Gluck (himself an antiquities dealer) and Stead feebly raise the issue of
anachronism, judging Pope by todays standards, Pope the scholar who
long ago also in normative fashion sold antiquities (xix, 584; as for Payne,
he was fully blind to archaeological research and goals). This ploy is disin-
genuous and a feint. For it is not so much that Pope was a dealer, like others,
such as E. Herzfeld (Ghirshman is mentioned in this role, 584), but the piv-
otal fact that selling antiquities was the force that drove his career, that his
public scholarly activities, Institute, Exhibitions, and publications were wil-
fully employed to screen and serve this dealing. Further, that he was both
aware of and utterly indifferent to the destruction he caused, transmogri-
fying what he knew was going on in Iran, what his dealing entailed, into
archaeological activity, is evident by his covering and protecting these facets
of his life from the scholarly community, and his invective against those who
suspected or knew. There is nothing anachronistic about reacting to this
abhorrent behavior, about removing the transparent fig leaf.
An overview of the work of Gluck and Siver have neatly allowed us to
place Popes negative evaluation of Negahban into proper perspective: a
Purveyor of Persian Art was defending his bazaar shop, unaware that at the
same time he was also bestowing a badge of honor.
4 The editors also include other negative comments received in their biographical re-
search: see pp. 170, 541, 585; Ackermans nephew thought Pope was a congenital liar.
Bibliography
Gluck, J. and Nol Siver 1996: Surveyors of Persian Art California, Mazda Publishers.
Legrain, L. 1934: Luristan Bronzes in the University Museum, Philadelphia, University
Museum.
Muscarella, Oscar White 1979: Unexcavated Objects and Ancient Near Art: Addenda,
Malibu, Undena.
Neghaban, Ezat 1996: Marlik: The Complete Excavation Report, Philadelphia, Univer-
sity Museum.
chapter thirty
THE ANTIQUITIES TRADE AND THE DESTRUCTION

OF ANCIENT NEAR EASTERN CULTURES*
Debate and evaluation of the commercial trade and dispersal of antiqui-

ties begins with the comprehension that the vast majority of unexcavated
ancient objects in existence, whatever their cultural backgrounds, have been
plundered; archaeologists did not excavate them (Nagin 1986: 23; Koczka
1989: 196). Archaeologists designate excavated material as artifacts, and
those non-excavated as antiquities. Possessors of antiquities often claim
they had been excavated, but this term can be used only to designate an
archaeological activity. Disorder, pertinent to both legal and archaeological
matters, occurs when museums and antiquity dealers, and some arche-
ologists, refer to provenance to identify a site-source of an unexcavated
antiquity. But the terms provenance and provenience are distinct, inasmuch
as they designate two distinct loci and two different activities. Provenience
specifically designates the site where an artifact was excavated; provenance
identifies the current or past location of the antiquity: a collector, museum,
auction house, or dealers shop (Muscarella 1977a; and pace Brodie et al.
2000: 3). Collector and museum catalogues and exhibition labels, along with
auction house and dealer catalogues, sometimes furnish a deceptive claim
that the antiquity derived from a named site, but they neglect to name the
attribution informant: a dealer or a previous auction house sale (Muscarella
1977c: 7779; 2000a: 11, 14; Vitelli 1984: 153). A fairly small number of antiqui-
ties were indeed plundered and traded decades ago, sometimes legally (e.g.,
commercial excavations in Iran). But these activities have never ceased;
they continue relentlessly throughout the world. Thus every topic and judg-
ment discussed herein obtains for every ancient culture in the world, with-
out exception; the ancient Near East is but one example of a worldwide
situation.
* This chapter originally appeared as The Antiquities Trade and the Destruction of
Ancient Near Eastern Cultures, in A Companion to the Archaeology of the Ancient Near East,
ed. D.T. Pott (Oxford: Wiley-Blackwell, 2012), 107124.
838 chapter thirty
The plunder of sites encased within a mound (Persian tepe, Turkish h-

yk, Arabic tell) formed by successive settlement constructions, and burials
and tombs, has a long history. The archaeological record reveals that the
practice occurred throughout antiquity. Numerous tombs in Egypt were
plundered millennia ago, the most spectacular being that of Tutankhamun,
which was looted after the kings burial but soon thereafter resealed. Royal
tombs built within a contemporary, inhabited palace at Nimrud (Iraq) were
partially plundered and then repaired while the site was still inhabited.
And numerous tombs buried under mounds of earth, called tumuli, and
visible to all, were totally or partially destroyed in antiquity and thereafter.
Examples include Pazyryk in the Altai (but much was recovered), S Girdan
in northwestern Iran, Kerkenes Dag in Anatolia, where scores of tumuli
there have been obliterated, and the Sardis area in western Turkey, where
90 percent of the tumuli have been plundered (Luke and Kersel 2006: 185
186; Roosevelt and Luke 2006: 173187).
The prevalence of ancient tomb plundering across the centuries within
Near Eastern cultural regions is unknown. But a good number of burial sites
have been excavated in modern times. Examples include tombs at Nimrud,
Kish, and Ur (Iraq); Umm-el Marra (Syria); Alaa Hyk and Arslantepe
(Turkey); Susa, Hasanlu, Dinkha Tepe, Marlik, and many in Luristan (Iran);
and Tillya Tepe (Afghanistan), where two burials were looted but six were
excavated, containing thousands of artifacts of gold, silver, and ivory (Hie-
bert and Cambon 2008: 210293). Artifact contexts of undisturbed burials
are not merely of inestimable value for knowledge of the ancient cultures
involved; in some cases, they are our only source of cultural data. They
also vividly inform us of the information forever lost from the countless
plundered tombs.
Ancient plundering was presumably conducted both as desecration and
to acquire loot. Looting is the basis for all current plundering, evidenced by
the vast number of destroyed cemeteries throughout the Near East. These
activities increased in the 19th century, a result of the renewed interest in
antiquity and fueled by a fulfillment of social ambitions exemplified by the
increased collecting of antiquities by museums and private collectors every-
where (Meyer 1973: 4647, 191197). Egypt, the Holy Land, Greece, Cyprus,
and Italy were the primary victims, never vacating that status. In the 1870s
plundering occurred somewhere near the wide-ranging Oxus river, and a
quantity of gold and silver objects (including modern forgeries) labeled the
Oxus Treasure was acquired by the British Museum. But no archaeologist
can identify its find-spot(s) (Muscarella 2003a). In the late 19th century Luigi
Palma di Cesnola looted countless sites in Cyprus. He sold thousands of
the destruction of ancient near eastern cultures 839
objects, subsequently smuggling and selling thousands more to the newly

established Metropolitan Museum of Art (MMA) in New York, for which
deed he was appointed the museums first Director, thereby establishing
its continuous plunder-purchasing tradition, and also encouraging other
museums to acquire antiquities.
Plunder also existed at this time in Iranfor example, Hamadan, the
Median capital, violated in the 1890s. In the 1920s exploitation in Iran ex-
panded, initiated by the destructions of Luristan cemeteries, financed by
Iranian dealers prodded by their growing number of foreign customers.
Luristan continued to be plundered for decades, and thousands of its antiq-
uities have been purchased (Muscarella 1988: 112120, 136206). Thanks to
years of excavations by Louis Vanden Berghe, scores of intact tombs were
recovered, providing for the first time local cultural contexts. Only one Luris-
tan habitation site has been excavated: Surkh Dum, in the 1930s (Muscarella
1988: 115135). And contrary to the belief among some archaeologists, sites
in Iraq, a state with anti-plunder laws, were also being looted in the 1920s
and 1930s. Numerous Iraqi antiquities were smuggled for sale to Iran, a
state with no anti-plunder laws. Hence, for decades scholars accepted as
archaeological fact that Mesopotamian artifacts, some bearing royal inscrip-
tions, derived from Iranian sites. Such presumptions resulted in erroneous
historical interpretations of alleged ancient Mesopotamian contacts east
of the Zagros Mountains (Muscarella 2000a: 15, 8121 n. 36). Forgers of
provenience, they produced a concomitant forgery of history, generated
from scholar-dealer cooperation, which is not the only example (Muscarella
1977b: 162163; 1977c: 7778). Scholars have also attributed stray Luristan
antiquities encountered as deriving from Armenia, the Caucasus, Iraq, and
Anatolia. Luristan antiquities displayed in the Adana and Van Museums
in eastern Turkey were confiscated from Iranian smugglers. I also saw in
a Van jewelry shop a lion pin stolen and smuggled from Hasanlu in Iran
(Muscarella 1988: 112113, 115; 2000a: 214 n. 56). In the 1920s antiquities from
Iraq also began to reach the expanding markets in Europe and the United
Statese.g., purchases by the newly formed Oriental Institute at the Uni-
versity of Chicago, competing with other museums.
Plundering essentially ceased during World War II, but soon thereafter
recommenced extensively across the Near East. The prime cause was the
appearance of more Luristan material. More momentous was the sudden
appearance of exquisite, hitherto unknown antiquities purported by deal-
ers and archaeologists (e.g., Andr Godard and Roman Ghirshman) to have
been discovered in 1947 at Ziwiye, in western Iran (Muscarella 1977a; 1988:
342349). Museums and collectors all over the world soon thereafter
840 chapter thirty
purchased them, and this continued for years as more Ziwiye material
surfaced. A number of the bronze, gold, and silver objects had been cut
into pieces and partitioned among the plunderers, an action resulting in
scores of fragments sold all over the worlda common practice (Hiebert
and Cambon 2008: 6779). The partition required years of work by scholars
to sort out and match the scattered fragments. Moreover, it was impossible
to know how many of the hundreds of artifacts purported to have come from
Ziwiye were actually recovered there or in fact came from elsewhere (other
sites, e.g. Qaplantu, have been proffered by dealers). Excavations at Ziwiye
by American and Iranian archaeologists recovered not a single comparable
artifact: but an historically important Urartian 7th century bc seal was
excavated there. The Ziwiye episode epitomizes the utter destruction of
a complex politys integrity and culture, and led to increased plundering
across Iran. Thus, following excavation in the southwest Caspian region at
Marlik, sites in the area were subsequently attacked. The Kalmakarra Cave
in Luristan in the early 1990s yielded scores of hitherto stylistically unknown
artifacts that have surfaced in the antiquities market (Muscarella 2000b: 30
n. 6). Another Iranian polity destroyed is that of the Sasanian kingdom in the
south. Hundreds of gold and silver artifacts labeled Sasanian have surfaced
over the years, many of which are genuine, but far more of those exhibited
in museums or collectors homes are modern forgeries (Muscarella 2000a:
203204, 528535 nn. 6870). Only some Sasanian sites in Iran and Iraq and
rock carvings have survived.
Plundering in Turkey also increased in the 1930s, accelerating after World
War II (Meyer 1973: 5764). The earlier examples include Aemhyk, Ho-
roztepe, and, allegedly, Ordu (Muscarella 1988: 394411). More recent cases
include Perge (Rose and Acar 1996: 7778; Renfrew 2000: 3234); possible
Hittite royal tombs, not one of which has ever been recovered; Elmali (Grae-
pler and Mazzei 1994: 92; Rose and Acar 1996: 8082), and the dynamiting-
mutilation of Phrygian rock-cut faades north of Afyon (which I have seen).
The 1965 destruction of several tumuli at Usak, east of Sardis, is a woeful
example of tumuli plunder (zgen and ztrk 1996; Greenfield 2007: 420
423; Waxman 2008: 135137, 144154). Here, hundreds of Greek, Lydian, and
Achaemenian artifacts, including painted frescoes torn from tomb cham-
ber walls, were soon thereafter purchased by the MMA, fully knowing their
geographical origin. The museum immediately assumed the role of their
guardian by hiding them in a storeroom for decades (I was secretly allowed
to see them). A small number were revealed in 1970, all deceitfully labeled
as Greek antiquities, a number of which were not. Not until 1993 did all
become public, when, after years of legal costs of millions of US taxpayers
dollars, the museum trustees admitted to having purchased antiquities they

knew derived from Lydia (according to minutes of the MMAs Acquisitions
Committee; Gross 2010: 356, 443, 445), and returned them to Turkey, their
homeland (Kaye and Main 1995: 150161; zgen and ztrk 1996; Rose and
Acar 1996: 7277; Muscarella 2009a: 399; 2009b: 1516). In all, 39 Phrygian
tumuli were excavated at Gordion in central Anatolia, approximately nine
of which had been plundered (at least three in recent times); about 10 were
excavated at Ankara and later at Bayindir, all intact. These tumuli deposi-
tions are our principal source of knowledge of Phrygian material culture.
Had the majority been plundered, the culture of a foremost Anatolian polity,
that of King Midas, would have been irrevocably eliminated, except for the
architecture excavated at Gordion and Bogazky.
Scores of Urartian cemeteries across northeastern Anatolia have been
destroyed. The irrevocable loss of information about the kingdom of Urartu,
a prominent polity, can be shown by the fact that museums and private col-
lections all over the world possess countless unexcavated Urartian objects;
excavated artifacts represent but a fraction of Urartian objects available
for study. One is confronted with articles and museum and exhibition cat-
alogues written by museum curators who readily inform us of a corpus
of antiquities in their possession that was found as a Fundkomplex (find
complex), Grabkomplex (grave complex), zusammenhngender Fund (asso-
ciated find), Sammelfund (hoard), einen seltenen Glcksfall (a rare piece
of luck) from Urartu, generously providing archaeologists with named sites
where the antiquities were (fortuitously) found. And the finds were curat-
ed, judiciously kept together by the finders and then by the smugglers. All
these data were forged in European curators offices, based on local deal-
ers claims (Muscarella 2006: 146, 213 n. 54). Almost 300 Urartian objects are
inscribed, but the plundered ones are orphaned historical documents, their
original sites unknown. The history and culture of Urartu have been irrevo-
cably crippled.
Plundering in Iraq has increased since the 1990s (Brodie and Renfrew
2005: 346; Muscarella 2007: 609; Brodie 2008: 6365). Five Iraqi museums
were looted in 1991 by locals during the reign of Saddam Hussein, their
contents scattered abroad (Bogdanos 2005: 491 n. 46; Muscarella in press
a). A surge occurred in 2003, which then slowed down (Brodie 2008: 69
71) but never ended. The Baghdad Museum was looted by its own personnel
(Bogdanos 2005, 2008b; see below) and because of irregular museum record-
keeping and missing files, most artifacts smuggled abroad cannot be traced.
This will cause legal problems regarding repatriation in the future. Since
the early 2000s, many Iraqi objects have been confiscated in Syria, Lebanon,
Turkey, the United States, and Europe.
842 chapter thirty
Pakistan and Afghanistan equally experienced devastating destruction of

their ancient cultures and history. Many sites in Afghanistan have been oblit-
erated and countless extraordinary antiquities were sold abroad. The Rus-
sian archaeologist Victor I. Sarianidi struggled for years excavating several
sites, enduring onerous conditions from local discord and thievesfueled
by their foreign customersponsors. Sarianidis excavations yielded architec-
tural, artifactual, and cultural information of a highly developed society of
the 3rd and 2nd millennia bc, a polity equal to that of the Sumerians. The
Kabul Museum was looted several times during insurrections in the early
1990s, but the museums staff acted courageously and saved many of its arti-
facts (Hiebert and Cambon 2008: 3541).
It is impossible to know how many artifacts have been taken from count-
less sites in the Near East since the 1930s, but there are unaccountable thou-
sands including an unknown number that remain hidden in museum and
dealer storerooms, awaiting a propitious time to be exhibited or revealed
for sale. Abetting this situation is that national and international so-called
anti-plunder laws are mostly disregarded or challenged. The often-praised
1970 UNESCO Convention on the Means of Prohibiting and Preventing
the Illicit Import, Export and Transfer of Ownership of Cultural Property
(Simpson 1997b: 297301) has, in fact, accomplished little to stop plun-
der. Most nations ignore it and difficulties (purposefully written into the
Convention) exist in its enforcement (Muscarella 1976; 2007: 603605, 616
n. 4; Elia 2000: 85; Gerstenblith 2006: 7680; Prott 2006: 3141; Vitelli and
Colwell-Chanthaphonh 2006: 56; Greenfield 2007: 214237). Furthermore,
and rarely noted, self-righteous museum proclamations contra plunder are
inadequate, or are even consciously ignored from the start (Muscarella 2007:
611614; 2009a: 403404). Only private treaties between nations or legal pro-
ceedings based on misrepresentations on custom declarations have some
effect (Vitelli 1984: 148150; Mallampati 2005: 120122; Gerstenblith 2006: 70
73,8283).
The objective of plunder is the acquisition of treasure to be sold: no cus-
tomers, no plunder. Universally, it is conducted by gangs of looters (often
known by the Italian term tombaroli) who work as organized teams. Often
they commit violence to defend their sites (Brodie et al. 2000: 1517; Daily
Telegraph, October 3, 2004). Dealers and their customers disingenuously
allege that their antiquities were merely found in the ground, that it was
a poor farmer plowing his field who accidentally made a chance find
(Atwood 2004: 288 n. 32). Or, in J. Cunos classic museum-speak/classical
critical theory-speak (to let readers know he has browsed Adorno): Its out
of the ground. Its out of the country. Its on the marketthe ground being
a nationalists buried cemetery, a tomb, or a mound (Bator 1982: 303306;

Elia 1997: 92; Mackenzie 2005: 5560, 213216, 229; Muscarella 2007: 612; in
press a). Sometimes the ground is a museum in Baghdad, Kabul, or Corinth
(looted in 1990; see Archaeology Online, February 6, 2001). Chance finds by
local peasants do occur; some are isolated examples, but others lead to mass
looting. Aside from Luristan, one egregious example is the plunder in 2001
of a number of cemeteries exposed by flooding to the south of Jiroft, in
southeastern Iran. Locals discovered intact burials filled with artifacts and
immediately began, not accidentally, to seek out others, selling their finds
to eager, indeed rapacious, dealers. Simultaneously, forgeries were manu-
factured and sold alongside the genuine loot, all labeled as from Jiroft
(Muscarella 2001). Subsequent archaeological activity in the area neglected
to investigate these cemeteries, to find out, as Vanden Berghe did in Luristan,
whether some burials had been missed: this was a serious archaeological
blunder. And whether the removal of hitherto unknown artifacts from the
Kalmakarra Cave resulted from a casual or a loot-seeking activity eludes
us. A common situation has existed for decades in the United States, Eng-
land, and Europe, where individuals become part- or full-time pot hunters
(Brodie et al. 2000: 2021), looting antiquities for their personal collections
or for sale to dealers and museums that openly encourage such activity. Pot
hunters are tombaroli. A different form of theft occurs, one difficult to iden-
tify, when local diggers at an archaeological site steal artifacts. The terracotta
tablet recording Sargons 8th campaign into Iran was stolen at the site of
Assur in Iraq by a local worker, who sold it; eventually, in 1910, it was smug-
gled into France and sold to the Louvre. And this was not a lone incident:
I have seen an ivory carving from Hasanlu on sale in an antique shop in a
town near the site.
Organized tombaroli sell their loot directly to their dealer-employers who
then pass it to organized smugglers for transport abroad by land, sea, or air.
Smugglers include airline and shipping personnel, individual entrepreneurs
and travel agents, as well as diplomats who conceal loot in their uninspected
luggage (Majd 2003: 3134 nn. 6&7, 73; Mallampati 2005: 117; Greenfield 2007:
247). Sometimes a smuggled antiquity is disguised as a modern copy pur-
chased as a souvenir (Gerstenblith 2006: 7173). All these activities involve
patent criminal behavior (Bogdanos 2008a: 57). Initial smuggling destina-
tions include Lebanon or Syria, thence to Europe, often Geneva, once a
primary destination but now more difficult because of local controls, then
to European, Japanese, and United States dealers, all legal ports for pas-
sage of antiquities (Muscarella 2007: 606; Bogdanos 2008b: 128). A growing
smuggling tactic is to arrange for shipments to be sent to Australia or other
844 chapter thirty
distant foreign ports, where dealers might employ schemes that camouflage
the original derivation, thereby allowing them to claim that the antiquities
belonged to a collector in Australia or Hong Kong, etc.in other words, a
forged provenance.
Clarification of the expressions plunder/looting, theft, illicit, and spoils
of war are essential, inasmuch as each may reflect a different episode and
background or a combination of more than one. Determination of illicit
acts is but a legal component of the plunder problem, one of the many
concerns involved (Bator 1982). To determine if a past or contemporary
acquisition involved legal or illegal acts, each requiring distinct courses of
action, all antiquities must be evaluated with regard to their acquisition
histories. A paradigmatic example, although not per se a Near Eastern mat-
ter, is determining which terminology is suitable in the ongoing discussions
between England and Greece regarding the repatriation of the Elgin Mar-
bles removed from the Parthenon in 1801 by Lord Elgins agent Philip Hunt,
and purchased from Elgin by the British Museum in 1816 (thereafter, con-
tinuous removal of sculptures occurred, acquired by European museums;
see Waxman 2008: 81). A number of metopes and sculptures were taken,
based, it was argued, on the firman, a government authorization. Elgins
role in acquiring the marbles, and past and ongoing negotiations for their
return, are fully documented by St Clair (1998) and Greenfield (2007: 4196).
Based on contemporary standards at the time of the removal, the issue is
partially one of looting, given that unauthorized removal was involved, and
bribes were paid to local workers, and threats were used. What remains to be
resolved legally (leaving aside national and historical issues) is the distinc-
tion between the metopes illegally removed, contra the firman, and those
legally removed. This action is an issue of modern law and is of equal con-
cern to archaeologists and the public. The objects that Elgin removed legally
do not come into the category of plunder.
The modern, legal ownership of the Rosetta Stone, discovered in Egypt
by Napoleon, is an example germane to spoils of war issues, not to those of
plunder or theft. It was captured from the French by the British and taken
to England, and then given to the British Museum. Egypt has demanded its
return. In another example, Egypts demand for the return of the head of
Nefertiti, now in Berlin, is different. This was certainly acquired by theft and
thus constitutes a modern legal matter: the head was stolen from Amarna
by the German archaeologist Ludwig Borchardt, who excavated it in 1912.
Borchardt deceptively hid the head in a crate beneath layers of sherds for
export to Berlin, subverting the legal division regulations for excavated arti-
facts. The Nefertiti head was kidnapped and smuggled into Germany: an act
accomplished by an archaeologist. It was a crate of sherds, not the Nefer-

titi head, as claimed by the Berlin Museum, that had Egyptian approval for
legal export (personal information from Rudolf Anthes in 1956). In Anthes
publication on the subject (1954: 19), he ducked the issue, disingenuously
asserting: The unique quality of the NofreEte [sic] head was apparently
not sufficiently stressed by those concerned (i.e. Borchardt; my italics). But
the concerned Egyptians knew nothing of the heads existence until Bor-
chardt in 1923 revealed its presence in Berlin. Removal of excavated material
from Iraq in the mid-19th century is another matter. For example, Austin
Layard excavated at Nimrud, uncovering hundreds of stone Assyrian reliefs
and statues. Some of these he donated to American colleges, others he
took home to England, where they were eventually sold to a dealer, and
thence to the MMA. By contemporary standards, these acquisitions were not
plunder.
The removal of excavated artifacts from Troy in the 1870s by Heinrich
Schliemann was also clearly a theft, as the objectsknown as Priams Trea-
surewere removed and smuggled abroad contra his firman. Negotiations
over the return of the treasure to Turkey has been ongoing for decades,
complicated by their present museum locations (see Simpson 1997a). In
1926, James Breasted purchased a gold tablet that had been stolen from
the Vorderasiatisches Museum in Berlin. A more recent staff theft occurred
within the Turkish museum in Antalya (Waxman 2008: 152154,162172).
Years ago, I was shown a cylinder seal donated to the Royal Ontario Museum
by a donor of objects. I asked a scholar to research it, and she discovered
that Leonard Woolley had excavated it at Ur, showing me its publication
photograph. The donor protested; he had purchased it from an honest Iraqi
sheikh who told him it had been found at site X. Once convinced, he returned
the seal to the Baghdad Museum where one of its staff had stolen and sold
it.
Antiquity dealers are the penultimate destination for plunder. They bear
sophisticated names such as Ariadne Galleries, Royal-Athena Galleries,
Phoenix Ancient Art, or simply the dealers name. They describe them-
selves as experts, esteemed, honest, and reputable. They save and sell art
acquired through trade and in good faith, implying legitimate acquisi-
tion (Muscarella 1977b: 159160; 2000a: 2; Koczka 1989: 190191; Atwood 2004:
31). The final markets for dealers antiquities are collectors and museums
(Kersel 2006; Muscarella 2007: 611614; 2009a: 404405; ini press a). Auc-
tion houses are also major vendors of antiquities, many of which have been
recently plundered (Brodie et al. 2000: 23, 2629). To disguise these antiq-
uities, dealers and auction houses provide a camouflage ruse, proffering a
846 chapter thirty
deceptive provenance by claiming that their antiquity derived from an old

private collection recently discovered in a basement in Italy or Germany, or
derived from a noble European family or from the Collection of Monsieur
R (Atwood 2004; Simpson 2005: 2930, 32; Muscarella 2007: 610; Christies,
London, 10/25/07: 83). To document a purportedly old provenance, dealers
will supply forged letters, eagerly embraced by their customers, as documen-
tation that the purchase was legitimate (Atwood 2004: 84). A prominent
example is the purchase by the Getty Museum of a Greek kouros in 1984,
for which acquisition a letter dated 1952 was presented to confirm that it
had belonged to a Swiss collector for decades, therefore acquired in good
faith. Some years later, the museums director announced that he had only
recently discovered that the letters envelope had an anachronistic postal
zip code and was a forgery (Elia 1997: 95; Lapatin 2000b: 4353; Renfrew
2000: 41); some scholars think the kouros itself is a forgery.
Dealers often cite an earlier auction sale as a provenance for their antiq-
uity, which is merely a record of yet another modern provenance. They
ship their antiquities for sale to foreign auction houses, enabling a pur-
chaser to claim a provenance in France or England. Dealers also utilize
auctions to sell their merchandise anonymously, especially when they sus-
pect it is a forgery. Another antiquity-selling market is the internetin par-
ticular, eBaywhere, alongside genuine artifacts, forgeries are offered for
sale (Stanish 2008). Countless postings offer objects alleged to derive from
Africa, Mexico, and South America (Kelker and Bruhns 2010). Such behavior
is classic bazaar archaeology (Muscarella 1995; 1999; 2003a: 264265; 2006:
151152, 157, 161165).
It is evident that museums worldwide have been and remain the fore-
most purchasers of plundered antiquities (Koczka 1989: 192193; Muscarella
2000a: 2325; 2007: 611612). Curators, some of whom are archaeologists,
initiate their museum acquisitions, seeking out and proposing purchases
(Muscarella 1974; 2007: 612613; 2009a: 400401; Cook 1995: 181, 185; Grae-
pler 2004), but ultimately directors and trustees make the final purchase
decisions. Unknown to most scholars and the public is that they make pur-
chases (and accept donations) knowing that they were plundered and smug-
gled abroad, an activity rarely reported in the press (for rare examples, see
E. Wyatt, The New York Times, 1/26/08: 1, 13; 1/30/08: All). Trustees include not
only wealthy and powerful citizens but also national and local government
officials and owners of important newspapers, all functioning in conflict-of-
interest roles (Silver 2006: 3; Muscarella 2009a: 399; 2009b: 7, 1112). Some
trustees collect antiquities, in part for eventual tax-deductible donations to
their museums (Nagin 1986: 24; Renfrew 2000: 2735; Atwood 2004: 141142;
Silver 2006: 1; Wald 2008; Muscarella in press b). Private collectors are also
wealthy individuals of social importance, exemplifying these roles by their
purchases. These are exhibited in, or donated to, museums, for which they
have galleries named after them, and receive tax-deduction benefits based
on the alleged increase in value since the original purchase (Brodie and Ren-
frew 2005: 353356; Silver 2006: 1,78; Greenfield 2007: 259). And noteworthy
is the fact that it is self-serving antiquities dealers who furnish the museum
appraisals. Collectors are cited by dealers and museum personnel as promi-
nent or serious (read serial) collectors, as having a lust or passion for
art, thus revealing their infatuation (Graepler and Mazzei 1994: 8184,87;
Muscarella 2000a: 9, 1113, 23 n. 5).
Consequently, pivotal to comprehending the nature of the plunder cul-
ture is full awareness that, worldwide, museums and private collectors are
the financers and sponsors, the beginning of the long chain of the process
(Muscarella 1974; Elia 1994: 20; Brodie and Renfrew: 2005: 349). An Iraqi
official addressing the value of antiquities succinctly articulated this in the
following words: For me, for you, it is all priceless, but for them [the plun-
derers] it is useless if they cant sell it in the market (S.L. Myers, The New
York Times, 06/26/02: 6).
Museums and collectors identify themselves as protectors of the worlds
culture, stewards of antiquities, Guardians of the Past, fulfilling a public
responsibility to collect (Muscarella 2000a: 129; Renfrew 2000: 30; Macken-
zie 2005: 158162; for an accurate elucidation of stewardship, see Lynott
and Wylie 2000: 3539). Curators will lie about the actually known site of
their museums antiquities, sabotaging archaeology, as well as mocking the
museums educational mission (e.g., the MMAs Usak purchase: see above).
To justify their deeds, museums and collectors identify plundered countries
in classic imperialistic language as source nations (Cuno 2008: 89); they
proclaim that plundered artifacts are merely the self-proclaimed cultural
property of these nations chauvinistic, nationalistic, indeed racist atti-
tudes (Cuno 2008: xxxiixxxv, 1315, 26, 124; Waxman 2008: 176). They insist
that antiquities are not the property of anyone nation, and to state other-
wise is a political construction, for they are the common property of a world
society, composed of encyclopaedic (read non-Near Eastern) museums
(Cuno 2008: 129, 139; Muscarella in press a); and plunderers (a word never
used by Cuno), sellers, and buyers are engaged in normal, licit business and
positive cultural transactions. Underlining this decree, which consistently
refrains from discussing how antiquities are obtained, is that once muse-
ums and collectors have acquired property from a source nation, ipso facto
it becomes their legal, non-racist, non-nationalistic property (Muscarella
848 chapter thirty
in press a). Government collaborators support these proclamations by for-

mulating laws favoring the import process (Muscarella 2007: 604605 n. 1).
Another strategy argues that nothing should be done to stop plunder be-
cause it would be like trying to stop drug smuggling, that restricting legiti-
mate dealers and customers from legally selling and purchasing antiquities
will promote a black market, notwithstanding the fact that the antiqui-
ties market is already a black market (Elia 1997: 87). If preventive measures
are established and enforced, both plunder and antiquity sales will dimin-
ish: no museum or serial collector could then exhibit or donate their illegal
purchases. Note also that source country is the term employed to define
the provenience of kidnapped women sold into prostitution slavery across
the world. Shes out of her home. Shes out of the country. Shes on the
market.
Professional archaeological behavior is an important component of this
review. A good number remain indifferent (Muscarella 2000a: 26n8; 2007;
2009a: 395396, 398405) or are publicly troubled solely within the areas
they excavate, fully ignoring others (Muscarella in press a). Some archae-
ologists remain unaware of the plunder culture and the contextual exis-
tence of unprovenienced antiquities possessing only modern provenances;
as students they were never informed by their professors (Muscarella 2000a:
910) and they pass down their lack of knowledge (but see Vitelli 1996).
Some fully ignore it. Further, nota bene, many university- and museum-
employed archaeologists actively support antiquity acquisitions. They col-
laborate with and advise dealers and collectors on their purchases (Mus-
carella 1977b: 160, 163164; 2000a: 38, 1315; 2009a: 398403 and n. 38; Vitelli
1984: 152154; G.G. Griffin 1989; Graepler and Mazzei 1994: 7374; Elia 2000:
85; Brodie 2008: 68) or write muted apologies for their roles (Muscarella
1980; Cook 1995). Archaeologists write articles and provide guidance for the
antiquity dealer-owned magazine Minerva (Muscarella 2009a: 403 n. 38; in
press a) and ones promoting antiquity collecting, such as Odyssey. Some
meet socially with dealers and collectors for collaborative purposes, provid-
ing them with advice, and give lectures on their excavations and research,
seeking prestige and financing (Muscarella 2000a: 2325 n. 5; 2007: 612614;
2009a: 401; in press b). Others accept employment with collectors, dealers,
and auction houses, recommended by their archaeologist professors.
Some archaeologists have also functioned as antiquity dealers them-
selves (Butcher and Gill 1993); others actively support them (Muscarella
2000a: 78, 23, 2526 nn. 7&8). Some have stolen artifacts from their sites
and sold or donated them to foreign museums and collectors (see above).
The most outstanding cases were Heinrich Schliemann (18221890; see
above), Roman Ghirshman (18951979), and Ernst Herzfeld (18791948).

Herzfeld was one of the most brilliant (and devious) Iranian archaeologists
known. He stole many artifacts that he himself had excavated at Persepolis
and other sites, then illegally, contrary to archaeological principles, smug-
gled them abroad via Swedish and German diplomatic luggage. He (and his
sister) then sold these on to several museums; he also sold forgeries (Majd
2003: 73, 197, 199, 200204; Muscarella 2005a). One example, looted from
Persepolis and now in the MMA, is a foot with an etched Greek drawing,
cut from a relief by Herzfeld and smuggled abroad (Muscarella 2005a: 431).
Nevertheless, scholars continue to defend Herzfelds crimes, claiming that
he was merely an avid collector of antiquities he collected small objects
(avid here being a synonym for lust and passion, thereby justifying rapists,
those of the earth and others: Muscarella 2000a: 12, 2324 n. 5); only men-
tioned in a footnote are his sales to the MMA of artifacts from the Persepolis
excavations (Mallampati 2005: 111112, 116; Muscarella 2005a). Such behav-
ior remains unfamiliar to most scholars, students, and the public. In the past,
archaeologists did sometimes purchase antiquities from dealers, which was
not considered a cultural crime at the time, and their collecting cannot be
judged by modern standardsa case in point is Andr Godard (18811965;
see Muscarella 1977a: 197; whether Godard sold antiquities is unknown).
Flinders Petrie purchased antiquities, but he did not steal artifacts from his
excavated sites; the same applies to James H. Breasted, who purchased mate-
rial for the Oriental Institute (Muscarella 2005a: 432).
Roman Ghirshman looted artifacts from his own sites and then gifted
them to foreign museums, which led to his being awarded Life Member-
ship of the MMA (in 1957); he also sold antiquities (Muscarella 2000a: 2526
n. 7). All the artifacts sold/donated by Herzfeld and Ghirshman were ille-
gally removed, thefts from Iran, their legal owner. As for Arthur Upham
Pope (18811969), he was one of the most powerful and duplicitous indi-
viduals involved in the destruction of Irans culture. He warrants discussion
both because he and others have asserted he was an archaeologist (Mallam-
pati 2005: 112), although he was not, and because he was for 45 years one of
the most active Iranian antiquity dealers known. Pope established archae-
ological organizations as scholarly fronts for his plundering activities, using
them as camouflaged archaeological venues for his dealer activities. He
commissioned thefts from Islamic shrines and purchased countless antiq-
uities, smuggling them abroad in diplomatic pouches. His writings defend
his archaeological responsibility to purchase and export antiquities, arguing
that forgeries (that he and others sold) were a minor collaborative problem
(Muscarella 2000a: 209211 nn. 36&38; 1999: 712; Majd 2003: 2953).
850 chapter thirty
Like Pope, Moshe Dayan (19151981) was not an archaeologist, although

he was lauded as a superb archaeologist and an amateur archaeologist,
labels he accepted. He was an Israeli General and later Minister of Defense.
As a General engaged in battle and continuing for decades, from 1951 to 1981,
he looted scores of sites in Israel, in contested state areas and the Egyptian
Sinai, and then sold the antiquities from his home to dealers, collectors, and
museums. Because of his power, the Israeli government and archaeologists
who opposed his activities could do nothing to stop him (Kletter 2003).
Recognition of these complex and intertwined areas is gradually increas-
ing. Beginning in the 1970s, some archaeologists began to write and lecture
about these issues, slowly joined by others; they became the incipient core of
professional opponents of plunder. Clemency Coggins was the first archae-
ologist to address the matter publicly, followed by Ezat Negahban (Meyer
1973: 3840; Muscarella 1999: 6); they converted the present author, who,
like most archaeologists, had no prior knowledge or appreciation of these
matters. Increasingly, more scholars (but alas, still a minority within the
discipline) have become actively involved, lecturing and writing on these
issuese.g., N. Brodie, C. Chippendale, R. Elia, B. Fagan, D. Gill, D. Grae-
pler, J. Greenfield, E. Herscher, M.M. Kersel, C. Renfrew, K.D. Vitelli, and
P. Barfordbut paradoxically, they rarely function collegially and do not
collaborate to organize sessions at professional meetings and conferences,
a disservice to students and the public. Linked to this issue is that most
professional archaeological organizations are only mildly active, do noth-
ing, or worse. For example, the Archaeological Institute of America (AIA)
functions as a venue for a small number of anti-plunder lectures delivered
at national meetings, mostly concerned with one specific country, and do
not promote lectures on plunder in their sponsored lecture tours across the
United States. More devastating, the AIA has vigorously embraced a plun-
derer, Indiana Jones, as a model for archaeology students; it has also joined
forces with active plunderers (Fagan 1996: 239; Muscarella 2009a: 398, 402; in
press b; SAFE Corner, 6/5/08). The Society for American Archaeology is the
single exception in the United States of a professional organization fighting
the fight (Lynott and Wiley 2000). And from its inception in 1974, The Jour-
nal of Field Archaeology, founded by archaeologists, functions as a singular
example of an archaeological journal actively engaged in plunder issues in
all areas of the world, its original aim. Web sites such as Culture Without Con-
text, Looting Matters, SAFE Corner, paul-barford and Jarvis (a compendium
of relevant publications) are also active in the fight, constantly publishing
reports from all over the world. Also, a growing number of publications
from scholarly conferences have appeared (e.g., Messenger 1989; Vitelli 1996;
Lynott and Wylie 2000; Brodie et al. 2001, 2006). A non-professional, lay orga-
nization in the United States, Saving Antiquities For Everyone (SAFE), has
been vigorously active in exposing and fighting the plunder culture, reveal-
ing its existence everywhere and naming those responsible.
Although under-appreciated by most scholars (and thus their students
too), the connection between the antiquities trade and the existence of forg-
eries is obvious. As the archaeologist E. Unger succinctly stated, As long
as there are people who collect antiquities, there will also be people who
forge antiquities (Muscarella 2000a: 12). Forgeries are created to be sold as
ancient antiquities. Virtually all museums and collectors worldwide have
purchased and labeled forgeries as ancient antiquities; several museums in
the United States possess nothing but forgeries. Archaeologists incorporate
forgeries into their courses, innocently integrating them into their teachings
of ancient cultures. Museum-employed curators and archaeologists exhibit
and publish forgeries, sometimes knowingly, obeying museum orders for
fear of offending rich collectors or their colleagues (Muscarella 1980: 117
n. 3; 2000a: 45, 1314, 3739). Archaeologists also, innocently or not, pub-
lish forgeries as ancient artifacts. A prominent example is the archaeolo-
gist Roman Ghirshman, whose many publications, both monographs and
exhibition catalogues, are standard scholarly texts. He was the most prolific
publisher of Iranian forgeries, baptizing them as ancient productions and
providing them with forged (by him) proveniences; many of them were in
the antiquities market (Muscarella 1977b: 164 n. 42a, 182 n. 83; 2000a: 28 n. 11,
34, 205 nn. 2&4). On one occasion, Ghirshman published details of antiqui-
ties he asserted had been recently discovered in Iran in the respected Illus-
trated London News (ILN, April 2, 1960: 550), then an important outlet for the
publication of actual archaeological discoveries. An unnamed archaeologist
(Ghirshman?) employed the same tactic in the same venue (December 1967:
5455) concerning two antiquities, alleging them to have been recovered
together with Parthian coins in Iran. On both occasions all the antiquities
presented were forgeries, their provenance a forgers workshop (Muscarella
2000a: 8). Since Ghirshman had published reports on his excavations in the
ILN, the ILN assumed this ruse was but another such report. Arthur Upham
Pope also deployed the ILN to sell his antiquities (Muscarella 2000a: 210
n. 38).
The detection of forgeries takes years of studying excavated artifacts, their
specific styles and motifs, as well as the structuring technologies and mate-
rials employed by individual cultures, i.e. to employ connoisseurship (a
word now condemned by some). Connoisseurship, like all heuristic inves-
tigations, is fallible but is absolutely essential for the study of artifacts and
852 chapter thirty
antiquities, with the caveat that, aside from scholarly mistakes and igno-
rance, it has its manipulators (Muscarella 1977b: 165169 n. 68; 1980: 118119;
Lw 1993: 3941; Simpson 2005: 2834; Grann 2010). While archaeologists are
becoming capable of recognizing forgeries, some (most?) who are anthropo-
logically trained are not. Brazenly rejecting these skills, they assert, Archae-
ology is anthropology or it is nothing and they scorn traditional archae-
ologists as object oriented self serving antiquarians (Muscarella 2000a:
1011; Wylie 2000: 139, 144). Accordingly, those who proclaim this off-the-wall
dogma theory ignore stylistic evaluations of artifacts they encounter (except
pottery), and lack both knowledge and interest in evaluating forgeries.
Some forgeries exhibit good workmanship and artistic skills (e.g., Wax-
man 2008: 153162), others reveal unskilled hands, incorporating stylistic
errors or anachronistic details (Lw 1993; 1998: 533562; Lawergren 2000;
Muscarella 2000a: 31215). Both categories are sold and published as ancient
objects. Forgers copy both forgeries and excavated artifacts. They also create
pastiches, utilizing a genuine core with the addition either of non-related
ancient or modern-made elements, or add engraved scenes to genuine
unadorned plaques or vessels (Muscarella 1999). Forgers often attempt to
create a unikum, a hitherto unrecorded type of artifact and therefore all the
more valuable to customers and scholars (Butcher and Gill 1993: 386; Mus-
carella 2000a: 1719, 209 n. 31; 1999; 2006: 166167). Forgers have lifetime jobs,
although some are now working part time.
Scholarly awareness of forgeries of ancient Near Eastern artifacts began
in the late 19th and early 20th centuries but then declined (Muscarella 1977b:
154155, 169 n. 68; 2000a: 9). Early discussions were primarily concerned
with a specific object or inscription, many of them alleged to be Hebrew
and Christian texts. A noteworthy example of such a forgery occurred in
Italy with the appearance of a gold fibula inscribed with the name Manios
that was presented by an archaeologist in 1887. The fibula was enthusias-
tically accepted because its inscription was considered to be the earliest
Latin writing ever recorded, a prize indeed. Years later it was revealed to be
a forgery, commissioned by the archaeologist, who was seeking professional
fame (Muscarella 2000a: 11; for a description of another gold fibula that even-
tually turned out to be a forgery or pastiche, see Simpson 2005: 2832). At
present, relatively few scholarly references to forgeries occur in archaeo-
logical literature, often through ignorance, but also because some scholars
deliberately suppress discussions to defend themselves or colleagues, which
deeds play a significant role regarding general ignorance of their existence
(Muscarella 1977b: 154156, 161163; 1980: 117118 n. 3; 2000a: 25, 710, 12; in
press a; Butcher and Gill 1993: 387, 396, 399 n. 4, 396 n. 36).
Thousands of forgeries of Ancient Near Eastern antiquities have been cre-

ated and sold in the post-World Ward I period. All antiquity dealers sell forg-
eries and some have collaborated with forgers for decades, especially those
in Iran. Some dealers who possess both forgeries and genuine antiquities
attentively offer them for sale separately, to suit their customers expertise
or lack thereof. Dealers advertise themselves as experts on forgeries and
state they easily recognize them; all assert they never sell forgeries. Cleverly,
they also donate forgeries to provincial museums, where levels of knowledge
amongst the staff are minimal, often dealing with the director who normally
is not an antiquity specialist, but who wants to build up the museums col-
lections. This gifting tactic is employed to win over the museums as future
customers and to stamp the gift as genuineif its in a museum, its real;
the dealer also claims a tax deduction. One unfortunate result of this is that
the Israeli government has created two postage stamps depicting forgeries
in the Israel Museum (Muscarella 2000a: 55, no. 33; 63, no. 11). Forgeries exist
of every conceivable type of ancient artifact and material. The materials
employed are extensive, but gold and silver are favorites because of their
inherent value and aesthetic appeal. Other materials include bronze, glass,
ivory, various stones, and terracotta (the latter are commonly forged in pre-
Columbian areas; see Kelker and Bruhns: 2010: 2021, 129160). Forgeries are
usually manufactured as unique, typical, or prominent antiquities, but for
tourists and less wealthy collectors there exist forgeries of simple, inexpen-
sive objectse.g., lamps, statuettes, carvings, seals, and objects with reli-
gious or erotic scenes (a bestseller).
Following a significant archaeological discovery or a recent plunder, forg-
ers immediately begin copying the excavated artifacts, a practice not limited
to the Near East (see Butcher and Gill 1993; Lapatin 2000a: 1828). Forged Ira-
nian antiquities are very common, resulting in countless examples based on
artifacts from Marlik, Luristan, the Achaemenian period, Ziwiye, Jiroft, and
Kalmakarra Cave (Muscarella 1977c: 7879; 2000a: 44133; 2001; 2003a). Since
the 1950s forgeries of Sasanian artifacts, in numbers exceeding those exca-
vated, have surfaced (Muscarella 2000a: 203205 nn. 6870, 528535). Forg-
ers often unwittingly produce stylistic or physical discrepanciese.g., eye
structureor misinterpret ancient manufacturing techniques (Lw 1993:
38; 1998: 525562; Muscarella 2000a: 31132, 206212 nn. 850; 2008a: 14; in
press b). From Mesopotamia come countless forged seals, heads and statues
of humans, deities, animals, vessels, etc. (Muscarella 2000a: 159187, 463
487). From Israel there are unique historical and religious inscriptions,
burial urns, tablets, and religious items (Silberman and Goren 2006: 49
63).
854 chapter thirty
Not surprisingly, Anatolia has experienced the same fate. Among the forg-
ers most successful productions are vessels and scores of nude terracotta
and stone female figurines alleged by dealers and their purchasers to have
been recovered at the Neolithic site of Hailar. These figurines are readily
distinguished from excavated examples because of crucial misunderstand-
ings both of the ancient manufacturing method and of the originals thigh
and leg positionsjoined, not separated (Muscarella 1980: 120; 2000a: 135
157, 434453). The use of stone for forged figurines has been described (accu-
rately!) by its museum purchaser as a Novum (Muscarella 1980: 120). A
number of plundered Urartian artifacts have been embellished with modern
engraved Urartian and non-Urartian decorative scenes, accepted by some
scholars and museums as data for enlarging our knowledge of ancient Urar-
tian art, history, and foreign contacts (Muscarella 2006: 153, 165166, 174). A
small number of Hittite forgeries exist (Muscarella 2000a: 143156, 454461).
A discovery reported by an archaeologist involves objects allegedly de-
rived from two tombs at Dorak, in northwestern Turkey. They were first
introduced in the ILN (November 28, 1959: 754, Pl. iiii) by the brilliant and
charismatic archaeologist James Mellaart. He later claimed that in 1958 he
met a woman on a train who took him to her home in Izmir, there show-
ing him objects from two tombs she said were discovered in 1922, which
is 36 years before Mellaart saw them. She allowed him only to sketch the
objects, one of which bore an Egyptian inscription, as well as the tomb plans
showing the objects in situ. Published in the ILN were painted drawings
created in Ankara from the sketches. If genuine, the corpus would attest to
the existence of a hitherto unknown, complex culture in western Turkey in
the 3rd millennium bc close to Troy. The precisely drawn tomb plans (first
published in 1967), with artifacts shown in place, explicitly suggested that
an archaeologist had painstakingly excavated them, and the drawn objects
are spectacular and unique. However, not a single object or photograph
has surfaced to date and many scholars believe that the Dorak Treasure
is a masterful (psychological?) fraud, the drawings and tomb plans modern
creations (Muscarella 1988; 397398 n. 4; 2000a: 141; Greenfield 2007: 416
417; S. Mazur in Scoop: 7/27/05, 10/4/05, 10/10/05 [www.scoop.co.nz]). The
very same problem applies to several drawings of fragmentary wall paint-
ings alleged by Mellaart to have been excavated at another of his sites, atal
Hyk. No archaeologist at the site had ever seen these fragments or the
paintings or their photographs (Muscarella 1988: 397398 n. 5; 2000a: 141
143). Forgeries implanted by archaeologists at their sites have also occurred
in Japan (Harvard Asia Quarterly, VI/3, 2002) and is alleged to have hap-
pened at a site in the United States (The New Yorker, 8/12/95: 6683).
Scholars and curators forge proveniences by stating that an unexcavated

antiquity, whether a forgery or genuine, is said to come from site X, neglect-
ing to mention that it was a dealer who said so (Muscarella 1977a: 216; 1977b:
163). Another deception, usually made by dealers, is to assert that there are
very few forgeries, that the quantity alleged is exaggerated and therefore
forgeries are a minor issuea claim made against the thousands in exis-
tence (Muscarella 2000a: 79; 2005b). Discovering that a forgery has been
purchased, curators become indignant: they have been deceived and dissim-
ulate to protect their own and the museums reputations. They will remove
the forgery from view or leave it on view, especially if it happens to belong
to a prominent collector whose antiquities are being exhibited (Muscarella
2000a: 24, 7, 9; Kelker and Bruhns 2010: 1214, 42, 5257). An archaeologist
curator at a Canadian museum accepted a donation he knew consisted of
forgeries so as to avoid offending the rich donor and losing future contribu-
tions (Muscarella 2009a: 401).
Unqualified scholars have accepted payments to write letters authenti-
cating forgeries, usually on their university stationery (Muscarella 2005b).
More commonly, conservators are employed by dealers and collectors to
authenticate their forgeries in writing. Written in positive terms, these
authentications often consciously avoid crucial structural and chemical
analyses (so as not to lose a good customer) and their positive reports
are subsequently defended as objective scientific endorsements (Elia 1995;
Tubb 1995a: 256260; Brodie et al. 2000: 18; 2002: 286290; Muscarella 2000a:
139140; 2008a: 1012, 1415; Grann 2010). Some conservators work indepen-
dently; others are employed by museums where they authenticate not only
museum acquisitions but also the personal purchases of the trustees (their
employers) and rich donors (Silver 2006: 35; Muscarella in press b). By
professionally performing these paid functions, they are partners and pro-
moters of a criminal plunder-laundering process. At least one honest inde-
pendent conservator has been banned from providing further consultation
by a United States museum because he reported that many of its possessions
from many geographical areas are forgeries. And furthermore, some con-
servators are themselves forgers, causing serious scholarly and legal issues
(Grann 2010).
To summarize fully the consequences of plunder would take more discus-
sion than is possible here. Concisely, then: it represents the partial or total
elimination of modern endeavors to acquire an accurate comprehension of
this planets ancient histories and cultures, to accurately situate the roots
and developments of modern civilization. This is the consequence of the
destruction of ancient tombs and civil and religious architecture that results
856 chapter thirty
in the eradication of original depositions and juxtaposition of artifacts. With

the original deposition locus unknown, their intended functions and mean-
ings are also unknown. Excavations require years of fieldwork, photography,
and drawings, and scores of years of close study, interpretation, and publica-
tion. Tombaroli need only a few days work in the ground. Consider the years
of in situ research required for the excavation and recording of the excavated
artifact depositions and funeral process of the tomb of Tutankhamun. Or the
countless months involved in on-site excavations followed by the recording
and drawing of the tombs architecture and the hundreds of juxtaposed arti-
facts found with the Phrygian king buried in Tumulus MM at Gordion (prob-
ably of King Midass father). It took Ezat Negahban 11 continuous months to
complete his excavations at Marlik (November 1961 through October 1962),
harassed continuously by thugs who attacked his camp, demanding the site
for themselves (Muscarella 2000c). The Iranian government had to send
police in to protect him.
Research on unexcavated antiquities permits scholars merely to study
mute, plundered antiquities only in a phenomenological sense, to attribute
them to a particular culture and date through connoisseurship. Because in
ancient times artifacts were sold, gifted, or dedicated to faraway centers,
archaeologists cannot attribute them to their depositional site, even if the
culture can be determined (Muscarella 1977a; 2000a: 1314; Elia 1997). There-
fore, unexcavated antiquities, along with (unexcavated) forgeries, both
attributed to ancient sites or cultures, create a fragmented and fictional his-
tory of the past. One example is the creation of Median art, a concept
derived solely from speculations on un excavated antiquities, both genuine
and forgeries, and for which polity not a single artifact is known to exist
(Muscarella 2000a: 46, 7375). All these matters articulate the dysfunction
within the archaeological realm.
In recent decades there have been a growing number of successful legal
suits by several nations to recover artifacts plundered from their ground
and sold to museums and collectors abroad (Gerstenblith 2006). In a num-
ber of cases, involving years of litigation, the plaintiff plunderers eventually
capitulated and returned the booty to its legal and natural source nation
(Prott 2006; Sokal 2006). The most prominent, successful cases have resulted
in the return to Italy of the MMAs Euphronious krater and artifacts from
Morgantina, and the same museums return to Turkey of the Usak plunder
(zgen and ztrk 1996: 1113). In the instances involving Italy, the museum
negotiated an exchange whereby vases excavated in Italy were loaned
(i.e., ransomed) for exhibition (Brodie and Renfrew 2005: 349350). The
MMA also returned a stone relief to Egypt. As a result of legal actions, some
objects from the Getty Museum, Boston Museum of Fine Art, Princeton Art
Museum, Cleveland Museum of Art, and the Louvre have been returned
to Italy, Greece, and elsewhere (Brodie et al. 2000: 9, 4748, 54; Waxman
2008: 6364, 298342, 356364). Also returned were a few plundered antiq-
uities from some collectors, including MMA Visiting Committee member
Michael Steinhardt (to Italy) and the collector and MMA trustee Shelby
White (to Italy and Greece). But for more than two decades, White consis-
tently refused to return the head of Hercules plundered from Perge in Turkey
(Rose and Acar 1996: 7778; Brodie at al. 2000: 32; Atwood 2004: 144145;
Waxman 2008: 137; Muscarella 2010), until September 2011, when she and
her co-owner partner, The Boston Museum of Fine Arts, returned the head
(a gift from White) to Turkey, but with the caveat that they had acquired it
in good faith (Boston Globe, September 11, 2011).
Guide to Further Reading
As primary and essential reading for scholars, students, and the concerned
public interested in the subject matters discussed herein, the volumes edi-
ted by Vitelli (1996), Brodie et al. (2000, 2001, 2006), Lynott and Wylie (2000),
Brodie and Tubb (2002), Brodie and Renfrew (2005), and Vitelli and Colwell-
Chanthaphonh (2006) are especially recommended, as are the publica-
tions of R. Elia, J. Greenfield, U. Lw, C. Renfrew, E. Simpson, K.E. Meyer,
S.R.M. Mackenzie, and O.W. Muscarella. Bibliographies in all these volumes
contain many more relevant works. Also strongly recommended are Atwood
(2004) and Watson and Todeschini (2006). Together, these last two books are
exemplars of the problem, documenting succinctly and fully the machina-
tions and worldwide connections of the plunderers and their facilitators.
For legal matters, a complex issue dealing with old and contemporary laws,
see Bator (1982), Simpson (1997b), St Clair (1998), Majd (2003), Gerstenblith
(2006), Prott (2006), and Sokal (2006). For scientific matters, Muscarella
(2008a) and Grann (2010) are recommended.
Bibliography
Anthes, R.J. (1954), The Head of Queen Nofretete, Verlag Gebr. Mann. Berlin.
Atwood, Roger (2004), Stealing History, St. Martins Press.
Bator, P.M. (1982), An Essay on the International Trade in Art, Stanford Law Review
34, 2: 275384.
Bogdanos, M. (2005), The Casualties of War: The Truth About the Iraqi Museum,
AJA 109: 477526.
858 chapter thirty
. (2008a), See No Evil, Museums, Art Collectors and the Black Markets They
Adore, in Rothfield 2008: 5761.
. (2008b), Thieves of Baghdad, in The Destruction of Cultural Heritage in Iraq,
P.G. Stone, J.F. Bajjaly, eds., The Boydell Press: 109134.
Brodie et al. (2000), Stealing History. Cambridge.
Brodie, N., et al. (2001), Trade in Illicit Antiquities: The Destruction of the Worlds
Archaeological Heritage. Cambridge.
Brodie, N. and Renfrew, C. (2005), Looting and the Worlds Archeological Heritage:
The Inadequate Response, ARA: 343361.
Brodie, N., et al. (2006), Archaeology, Cultural Heritage and the Antiquities Trade.
University of Florida.
Butcher, K., Gill, D.W.J. (1993), The Director, the Dealer, the Goddess and Her
Champions: The Acquisition of the Fitzwilliam Goddess, AJA 97: 383491.
Cook, B. (1995), The Trade in Antiquities: a Curators View, in Tubb 1995: 181192.
Cuno, J., (2008), Who Owns Antiquity? Princeton.
Elia, R.I. (1994), The world cannot afford many more collectors with a passion for
Antiquities, The Art Newspaper, No. 41, October: 1920.
. (1995), Conservators and Unprovenanced Objects: Preserving the Cultural
Heritage or Servicing the Antiquities Trade, in Tubb 1995: 244255.
. (1997), Looting, Collecting, and the Destruction of Archaeological Resour-
ces, Nonrenewable Resources 6, 2: 8598.
Gerstenblith, P. (2006), Recent Developments in the Legal Protection of Cultural
Heritage, in Brodie et al 2006: 6892.
Graepler, D., et al (1993), Fundort: Unbekannt. Munich.
Graepler, D. (2004), Archaologie und illegaler Antikenhandel: Die Rolle der Univer-
sitatssamlungen, in Illegale Archaologie? Weissensee Verlag: 116130.
Grann, D. (2010), The Mark of a Masterpiece, The New Yorker, July 12 & 16: 5071.
Greenfield, J. (2007), The Return of Cultural Property. Third Edition. Cambridge
University Press.
Hiebert, F. and Canibon, P., eds., (2008), Afghanistan: Hidden Treasures from the
National Museum, Kabul. National Geographic.
Jarvis, H., The Looting Question Bibliography. Internet: Documents/lootbib.
Kaye, L.M. and Main, C.T. (1995), The Saga of the Lydian Hoard antiquities: from
Usak to New York and back and some related observations on the law of cultural
repatriation, in Tubb, ed.: 150162.
Kelker, N.L. and Bruhns, K.O. (2010), Faking Ancient Mesoamerica. Left Coast Press.
Kletter, R. (2003), A Very General Archaeologist-Moshe Dayan and Israeli Archaeol-
ogy, Article 5 in The Journal of Hebrew Scriptures 4.
Koczka, C.S. (1989), The Need for Enforcing Regulations on the International Art,
Trade, in Messenger 1989: 185208.
Lapatin, K.D.S. (2000), Boy Gods, Bull Leapers, and Mother Goddesses, SOURCE XX,
1: 1828.
Low, U. (1993) Kunsthandel und Falschungsproblamatik, MDA-Ve v. Heft 1: 3641.
. (1998), Figurlich verzierte Metalalgefasse aus Nord- und Nordwestiran, Mun-
ster (reviewed by O.W. Muscarella in BibOr LVII, No. 1/2 188195).
Luke, C. and Kersel, M. (2006), Mysterious Shepherds and Hidden Treasure: The
Culture of Looting in Lydia, Western Turkey, JFA 31: 185198.
Lynott. M.I. and Wylie, A. (2000), Ethics in American Archeology. Society of American
Archeology.
Mackenzie. S.R.M. (2005), Going, Going, Gone: Regulating the Market in Illicit An-
tiques. Institute of Art and Law.
Majd, Mohammad Gholi (2003), The Great American Plunder of Persias Antiquities.
19251941. University Press of America, Inc.
Mallampati, H. (2005), Archaeology and Collecting: Law, Ethics, Politics, in M.C.
Root, ed., The Fertile Land Signs and Symbols in the Early Arts of Iran and Iraq.
Messenger, M. (1989), The Ethics of Collecting Cultural Property. University of New
Mexico Press (reviewed by O.W. Muscarella, AJA, 95, 1991: 342344).
Meyer. K.E. (1973), The Plundered Past. Atheneum.
Muscarella, O.W. (1973), Antiquities and Collections: A Curators Viewpoint, JFA 1:
25.
. (1974), The Antiquities Market: A Colloquium, Paper Two, JFA I: 221222.
. (1976). Antiquities Legislation Pending in Congress, Archaeology, 29, 275
276.
. (1977a), Ziwiye and Ziwiye: The Forgery of a Provenience, JFA 4: 197219.
. (1977b), Unexcavated Objects and Ancient Near Eastern Art, in Mountains
and Lowlands Essays in the Archaeology of Greater Mesopotamia, L. Levine, T.C.
Young, eds. Undena Publications: 153205.
. (1988), Bronze and Iron, Ancient Metal Artifacts in the Metropolitan Museum
of Art, New York.
. (1980), Die gefalschte Kunstgechichte-Ein Hausputz in der Vorderasiatis-
chen Altertumskunde: A Reply, APA 11/12 117120.
. (1995), Bazaar Archaeology, in Beitrage zur Kultgeschichte Vorderasiens,
Festschrift for Michael Boehmer. Mainz: 449453.
. (2000a), The Lie Became Great. Styx.
. (2000b), Editor, Special Issue on Forgeries of Ancient Art, SOURCE, XX, 1.
. (2000c), Excavated in the Bazaar: Ashurbanipals Beaker, in Muscarella
2000b: 2937.
. (2000d), The Pope and the Bitter Fanatic, in The Iranian World, Essays on
Iranian Art and Archaeology, eds. A. Alizadeh et al, Iran University Press: 513.
. (2001), Jiroftand Jiroft-Aratta, BA1 15: 173198.
. (2003), Museum Constructions of the Oxus Treasure: Forgeries of Prove-
nience and Ancient Culture, ACSS 9394: 259275.
. (2005a), Review of Ernst Herzfeld and the Development of Near Eastern Stud-
ies, 19001959. A. Gunter and S. Hauser, eds. Leiden. 200S, JAOS, 125: 431432.
. (2005b), Some thoughts on fakes, addendum, N.A.B.U. no. 4: 9798.
. (2006), Urartian Metal Artifacts: An Archaeological review. ACSS 12: 147175.
. (2007), Archaeology and the Plunder Culture, IJCT: 602618.
. (2008), The Veracity of Scientific Testing by Conservators, in Original-Copy-
Fake?, E. Pernicka, Silke von Berswordt-Wallrabe, eds. Mainz am Rhein: 918.
. (2009a), The Fifth Column in the Archaeological Realm: The Great Divide,
in Studies in Honor of Altan Cilingiroglu, eds. E. Gens, D. Celik. Istanbul: 395406.
. (2009b), A Review and Addenda: Michael Gross Rogues Gallery, SCOOP,
Internet, September 16, 2009: 117.
. (2010a), Archaeologists and Acquisitionists, IJCT 18, 2: 449463.
860 chapter thirty
. (2010b), An Unholy Quartet: Museum Trustees, Antiquity Dealers, Scientific

Experts, and Government Agents, in Memorial Volume for Massoud Azarnoush;
in press.
. (2010c), Shelby Whites Foundation Expansion, SAFE Corner (Internet),
June 1.
Nagin, C. (1986), Patrons of Plunder, Boston Review, August: 56, 2325.
zgen, L. and ztrk, J. (1996), Heritage Recovered The Lydian Treasure. Ankara.
Renfrew, C. (2000), Loot, Legitimacy and Ownership: The Ethical Crises in Archaeol-
ogy. Duckworth.
. (2006), Museum Acquisitions: Responsibilities for the Illicit Traffic in Antiq-
uities, in Brodie 2006: 245257.
Rothfield, 1., ed., (2008), Antiquities under Siege. Altimara Press.
Silver, V. (2006), The role of museum trustees, Bloomberg News, February 25: 113.
Simpson, E., ed. (1997), The Spoils of War, Harry M. Abrams.
Simpson, E. (2005), Tall Tales: Celts, Connoisseurs, and The Fabrication of Archae-
ological Context, SOURCE XXIV, 2: 2841.
St. Clair, W. (1998), Lord Elgin and the Marbles. Oxford.
Stanish, C. (2008), Forging Ahead, Archaeology (May/June): 18: 5866.
Tubb, K.W. (1995), The Antiquities Trade: an archaeological conservators perspec-
tive, in Antiquities Trade or Betrayed Legal, ethical and conservation issues, Kath-
ryn Walker Tubb, ed. Archetype Publication: 256263.
. (2002), Point, Counterpoint, in Brodie, N. and Tubb, K.W., eds., Illicit Antiq-
uities, The Theft of culture and the extinction of archaeology. Routledge: 286292.
Vitelli, K.D. (1984), The International Traffic in Antiquities: Archaeological Ethics
and the Archaeologists Responsibility, in E. 1. Green, Ethics and Values in Archae-
ology. The Free Press.
Vitelli, K.D., ed. (1996), Archaeological Ethics. Altimara Press.
Vitelli, K.D. and C. Colwell-Chanthaphonh, eds. (2006), Archaeological Ethics. Sec-
ond Edition Altimara Press.
Waxman, S., Loot (2008). Times Books. New York.
chapter thirty-one
THE FIFTH COLUMN WITHIN THE ARCHAEOLOGICAL REALM:

THE GREAT DIVIDE*
Abstract: The issues discussed here have been on my mind for a long time, and
some have surfaced in earlier publications.1 Here I expand on them, believing they
are appropriate for a Festschrift for Altan ilingiroglu, an excellent archaeologist, a
fine and honest scbolar and teacher; and a dear friend with whom I excavated for
seven years at Ayanis on Lake Van. In his honor I address a significant archaeological
condition, one regularly avoided or covered up: the contra-archaeological behavior
of some archaeologists. Their professional and moral behavior has resulted in the
multi-faceted, fractured, and unanchored nature of much of the discipline today.
The Plunder Culture
Most archaeologists are aware of the on-going plunder of sites and concomi-
tant destruction of this planets history, yet relatively few are actively con-
cerned with terminating the destructions. Gradually more and more infor-
mation is being published by archaeologists, and by non-professionals in the
press or the Internet and in investigative reports and books, thus increas-
ing attention to the topic. Archaeologists who over the years have consis-
tently fought the fight include C. Coggins, E. Negahbanthese two were
the first archaeologists to call attention to the problem (they converted me),
K.D. Vitelli, N. Brodie, R. Elia, E. Herscher, C. Renfrew, D. Graepler, D. Gill,
and C. Chippendale. And the number of archaeologists who undertake sim-
ilar actions in classrooms is increasing. Nevertheless, many archaeologists
remain reticent about engaging in vigorous and long-term efforts to stop
plundering and collecting activities that are contrary to the nature of their
discipline. Their indifference is matched by archaeological organizations in
Europe and the United Statesviz. the Archeological Institute of America:
manifested by the paucity of sponsored lectures and on-going professional
and public discussions; the absence of active lobbying in the press and with
* This chapter originally appeared as The Fifth Column within the Archaeological
Realm: The Great Divide, in Studies in Honour of Altan ilingiroglu, eds. H. Saglamtimur,
Z. Derin, and E. Abay (Istanbul: Arkeoloji ve Sanat Yaynlar, 2009), 395406.
1 viz. Muscarella 1991; 2000: 2229; 2008a: 613614.
862 chapter thirty-one
government agencies to stop plundering; and then their Presidents recent

pronouncement2 that the movie character Indian Jones has played a signif-
icant role in stimulating the publics interest in archaeological exploration,
when any first year archaeology student knows that Jones is the very antithe-
sis of an archaeologist. In fact, he has played a significant role in stimulating
the destroyers of sites, the plunderers who supply antiquities to a museum.
Once every few years a resolution to oppose plunder is unanimously
voted at annual meetings: all those against plunder, raise your hands, now
lets listen to lectures and socialize. Archaeological organizations period-
ically offer session on plundering at annual meetings, but rarely proceed
beyond revealing the problem, and concentrate on the speakers area of spe-
cialization. Unaddressed are long-term strategies in which they should be
engaged.3 And seldom are attempts made by the small active anti-plunder
archaeologists to maintain on-going contacts and coordinate actions; and
the same names continuously appear. What is common are professors who
occasionally mouth pious platitudes on plunder in their classrooms, or drop
brief comments in print or in lectures that it is not nice to buy plundered
antiquities, acts sufficient to demonstrate how archaeologically correct they
are. Some may be quite vocal4but not quite active. This is why most peo-
ple remain unaware that plunder and cultural destructions are sponsored
and directed by a financially, culturally, and politically powerful interrelated
alliance.
Citizens of the Plunder Culture are active throughout the antiquities pro-
ducing and antiquities acquiring world, functioning in interlocking levels
as mutually supporting aggressive columns attacking archaeology. At the
lowest level are the on-site, professional plunderers of sites and tombs, those
who carry out their customers orders. In Plunder Culture literature they are
always labeled poor farmers, who non-stop accidentally discover antiq-
uities while plowing their fields.5 The word tombaroli, used to describe
organized Italian plunderers, can apply to all. Above this level are the local
criminal and Mafioso column, consisting of local dealers, smugglers, corrupt
officials, who facilitate the transfer of the plunder abroad where they are
laundered before being shipped to international market destinations. The
third column consists of antiquities dealers (all referring to themselves as
2 Internet, Agade News, 5/16/08.

3 Also see BrodieRenfrew 2005: 357358.
4 Winter 1992: 32.
5 One example is Griffin 1989: 109. For a report on the profit gained by these plowers in
Jordan, see RoseBurke 2004.

the fifth column within the archaeological realm 863
reputable), who have standing orders from the level above to supply them
with antiquities, and forged documents manifesting tbe antiquity derived
from an Italian or German family that possessed it for generations.
The dealers sell their plunder to the upper level, the fourth column. It
consists of wealthy private collectors, university, private and public museum
directors and their curatorial staffs and trustees, and is supported by news-
paper owners and by elected government officials. Plundered artifacts are
referred to as antiquities, or Art.6
Collectively, Plunder Culture members are the raison d etre for looting
and destruction. The upper membership level, museums and collectors,
purchase antiquities as an on going class-power-identity activity. They ini-
tiate the process by sponsoring and financing the purchase orders, which
pay the on-site looters to destroy (the power, the power!) ancient sites, the
smugglers, the bribes, so that they can possess/save Art. For museum cura-
tors, some are archaeologists, others art bistorians see Muscarella 2000: 23,
13, 2325, 27, footnotes 5, 7; the acquisition of antiquities is a major compo-
nent of their job description, for which raises and promotions reward them. I
once heard one of the very rich, and thereby very powerful curator-sponsors
of plunder, Dietrich von Bothmer of the Metropolitan Museum of Art, sneer
at corrupt Italian officialsall lower class of course, for accepting bribes:
which he himself had paid!7 Courtesy of their social and financial status,
upper-level members also have the power to conceal their roles effortlessly,
for example keeping their activities out of newspapers and other media (this
is gradually changing).
Although the Plunder Cultures activities are becoming better known, pri-
marily from non-archaeologist writers and reporters,8 and increasingly from
professional archaeologists, the reports rarely present full details about the
wealthy and upper-social class perpetrators. A recent volume with essays
by scholars on plunder by scholars is Brodie et al. 2006;9 earlier examples
6 For newspaper owners and staffs suppressing information about plundering museums,
see Meyer 1974: 11, Horsley 1997, Atwood 2007 (also below), and Muscarella in Mazur 2006: 2
3.
7 In an Internet interview I referred to DvB as a son of a bitch. I wish to make an apology
for that stupid remark. I have been blessed by living with many female dogs, all loving,
truthful, and loyal bitches, and I apologize to them all for my solecism.
8 Two of the best recent books on this issue are works of Atwood 2004, and Watson
2006 (see also Watson 1977); also Nagin 1986. To date the only substantial public reports
on the archaeological issues raised here occur on the Internet: see Mazur 2006. A work that
presentsin a long and rambling mannerthe arguments and positions of dealers, auction
houses, museums, and legal issues (but little on archaeologists positions), is Mackenzie 1988.
9 Brodie et al. 2006; see also Muscarella 2008a.
of collected articles on plunder include Brodie et al. 2000 and 2001. David
Gills internet blog is an important example of professional engagement;
equally the McDonald Institute for Archaeological Research at Cambridge
University (now non-functioning), via its journal Culture Without Context
and monographs.10 In the past the Journal of Field Archaeology had a unique
Antiquities Market section edited by K.D. Vitelli, and later Ellen Herscher;
recently it has been revived. Very important is the public organization, Sav-
ing Antiquities For Everyone (SAFE), founded by lay people (Cindy Cho and
Paul Kunkel), ordinary citizens. They do work that not many archaeologists
and their organizations attempt: bring individuals of lay and professional
backgrounds together, to monitor, expose and fight plunder and those who
entail it, wherever encountered.11 Maybe eventually (insallah) professionals
will wake up and use them as a model.12
The Fifth Column
Another noteworthy and powerful component of the Plunder Culture exists

but remains unrecognized by most archaeologists and the public. In addi-
tion to the four visible columns attacking preservation of the worlds his-
tory, an active undercover Fifth Column operates within the archaeological
community and collaborates with the Plunder Culture citizens. I speak of
archaeologists who willingly work as partners with the upper (wealthier)
acquisition level. Two components exist. One, crucial, involves the ideo-
logical divide in the nature, methods, and goals of archaeological/academic
scholarship that exists between the professional behavior of (most)
museum-employed curators and that of (many; see below) university em-
ployed archaeologists.13 That most curators of ancient arta number of
whom have degrees in archaeologyfunction under entirely different
job descriptions, scholarly and methodological agendas and ideological
10 viz. Brodie et al. 2000 and Brodie et al. 2001.

11 Muscarella 2008a: 615.
12 I also want to single out Charles Koczka, a diligent and honest Customs agent, who
continuously fought to prevent illegal antiquities arriving in the United States. He rarely
succeeded, but he did play an important role in getting back to Turkey the Metropolitan
Museum of Arts plunder, The Lydian Treasure (see below; see also his article in Messenger
1989: 185198; Muscarella 1991: 344).
13 For discussions and documentation of this ideological/professional divide see for
example my articles: Muscarella 1977; 2003; 2006; also Simpson 2005. I disagree with Winters
1992: 31 challenge to what she calls the oppositional stereotypes such as museum curators
vs. academics.
loyalties is unknown to the discipline at large. This ignorance exists notwith-

standing that many curator/archaeologists sabotage archaeology in several
ways. I present but a few examples of the archaeology/museum divide, the
fifth-column sabotage or fellow traveler behavior involved.
I begin appropriately with a statement alleging the common goals of
archaeologists and museum curators proclaimed by a professor of archaeol-
ogy and President of the Archaeological Institute of America, C. Brian Rose.14
To him, the divide between these disciplines should be narrowed, because
both [italics mine] deplore the destruction of ancient sites, i.e., both are
colleagues in the battle against plundering (a letter written to Archaeology
challenging this was not published). The same belief in a common ground
between archaeologists, collectors and museum curators, that all are pas-
sionate in their common interests, is argued by Kathryn Walker Tubb.15
The rhetoric that plunderers and archaeologists share the same passion is
also presented by Griffin: Collectors and scholars must work together 16
(query: will the workshop venue be a dealers shop?).
As for Roses AIA Declaration, it is ironically juxtaposed toand thereby
exposes it as falsea photograph of the plundered Euphronios krater in the
Metropolitan Museum of Art: an antiquity purchased by museum staff:
who most surely did not deplore the destruction of the ancient Italian site
or otherswhence the vase derived. Roses declaration on what curators
deplore is the very opposite of what archaeologists deplore. To equate the
activities of the curators of ancient art of the Metropolitan Museum, the
Getty Museum, the Cleveland Museum of Art, The Boston Museum of Art,
the Los Angeles County Museum of Art, The Kimball Art Museum, The Saint
Louis Art Museum, the MIHO museum, the British Museum, the Louvre, and
many university museums, with archaeologists who painstakingly excavate
to save artifacts for everyone, is embarrassing.17
An example of a museum curator (Princeton) gushing about his activ-
ities is that of Griffin,18 who exults about his purchases of plundered pre-
Columbian artifactsboth for himself (he is also a private collector) and
his university. Initial collecting soon turned passionate, and Collecting, in
some miraculous way, has helped to create my life and its directions;19 only
14 Rose 2007: 6.
15 In Brodie 2006: 289290; see Muscarella 2008a: 615616 for discussion.
16 Griffin 1986: 465.
17 See examples below; also Muscarella 1991: 343.
18 Griffin 1986, 1989.
19 Griffin 1989: 106.
Purists would try to stop collecting for Collecting can be positive and
creative. Further, he is a serious collector.20 To him, it is archaeologists
not collectorswho are self righteous.21
For Harvards (one time) curator James Cunos sophisticated homage to
orphaned artifacts see Muscarella 2008a: 612. His museum purchases
182 Greek vessels in 1995 aloneprove he can deceive by transmogrifying
plundered artifacts and destroyed sites into Art belonging to everyone.
For similar beliefs we have a director of the St. Louis Museum in Missouri,
James Burke, claiming the world is a better place when world cultures are
shared 22 (also when the worlds women are shared in the same manner?).
Planned dissimulation was orchestrated by the Metropolitan Museum of
Art curator Dietrich von Bothmer, the museums director Le Comte Guy-
Philippe Lannes de Montebello, and Arthur Ochs Sulzberger, the owner of
the New York Times and (conflict-of interest) museum Trustee, serving for
decades on its Purchasing Committee. They purchased, exhibited and pub-
lished purchased artifacts plundered from Turkey, and baptized A Greek
and Roman Treasury.23 The Trustees and curatorial staff knew that the
objects had been plundered from now-destroyed tombs in western Turkey,
which was revealed many years later when the artifacts were returned to
their correct archaeological and geographical locus: in western Turkey, not
Greece.
The deceit has never been discussed openly within the museum, nor ever
been reported in the New York Times; only one employee in the museum
objected to the purchase.24 Karl Meyer obliquely reported the conflict of
interest behavior of Sulzberger when discussing the purchase of the plun-
dered Euphronios vase: for the paper to editorially reproach the Met would
involve criticism of the publishers friends.25 The conflict of interest role was
confirmed by Ken Auletta in The New Yorker (June 28, 1993: 4). He reported
that Sulzbergers son Arthur refused to become a Trustee at the Metropoli-
tan Museum in order to avoid any appearance of conflict of interest with
his duties as publisher. Another disturbing revelation of the behavior of
New York Times staff and editors is that of C.B. Horsley who documents
their involvement in suppressing of any information of the Metropolitan
20 Griffin 1989: 113, 114, read: serial collector.

21 Griffin 1986: 464465; see Muscarella 1989: 343.
22 In Nagin 1986: 6.
23 Muscarella 2008a: 607.
24 Muscarella 2000: 25, footnote 7.
25 Meyer 1974: 11.
Museums purchase of major forgeries of Chinese paintings.26 Roger Atwood

provided yet more detail about the duplicity and dishonesty of the Times
staff, confronting their conflict of interest positions in suppressing the antiq-
uities purchasing actions of Sulzberger and his position on his Museums
Acquisition Committee.27 There he sits alongside Shelby White; together
they purchased countless plundered artifacts.28
Years ago the Metropolitan Museum of Art hosted a lavish exhibition of
borrowed plundered objects loaned by a foreign, major plunder-sponsoring
museum accompanied by a sumptuous catalogue, all financed by the lend-
ing museum.29 The borrowing museums curators recognized that among
the plundered artifacts were a number of probable modern forgeriesan
opinion shared by me. Because museum employees obey a code that com-
mands that one never inform a rich patron or lender that their antiquities
are forgeries, the borrowing museums head curator decided to exhibit and
publish the forgeries as genuine ancient artifacts: casually deceiving schol-
ars, students and the public. This curator invited a European scholar to write
the catalogue entries for these forgeries and the genuine ones; a generous fee
would be paid. The scholar declined, noting that some of the objects were
forgeries. Following this un-museum behavior, the curator dispatched two
curators abroad to request the European scholar to reconsider the rejection.
Both emissaries are archaeologists, both knew that the challenged objects
were forgeries, both accepted the assignment. They were instructed to con-
vince the scholar to publish the forgeries as authentic artifacts, that the
word forgery could not be mentioned, but allowed the scholar to use the
term unparalleled when describing them. The recipient of this request to
lie, delivered by two archaeologist-museum curators, was upset and embar-
rassed by the professional insult, and rejected the offer. Another individual
was recruited to write the entries and readily complied.30 Curators, dealers,
auction houses, and collectors cite the reports of conservators to prove that
their purchased antiquity is a genuine plundered artifact.31
One of the above-mentioned emissaries recommended a plundered arti-
fact (not for the first time) for purchase to the museums Trustees (Sulzber-
ger, et al.). A colleague, concerned that the object had been plundered and
26 Horsley 1997.
27 Atwood 2007: 37.
28 See Muscarella on the Internet, December 12, 2005, February 9, 2006, via S. Mazur
(2006), also BrodieRenfrew 2005: 349350; on the MMA Trustees, see Nagin 1986: 6.
29 Muscarella 2000: 4.
30 For the objects see Muscarella 2008b: 1314.
31 For specific details of this still unrecognized problem see Muscarella 2008b.
torn from its now-destroyed locus, objected to its purchase. The retort was:
If you dont like what we do here, you can leave.32 The Trustees purchased
the plundered artifact. And the archaeologist-curator was awarded a profes-
sorship at a major university (recommended by archaeologists associated
with the awardees museum). The protesting curator was subsequently dis-
missed.
Witness also the 180-degree professional and moral turnaround by an
archaeologist who applied for the position of an art museum director. He
was appointed and immediately accepted his job description duty, to
actively purchase plundered artifacts, and employ all the museum-ritual
language to justify such actions.33 The metamorphosis from archaeologist
to Fifth Column citizenship was accomplished with ease. As a reward for
building up his art museums antiquities collection, he was subsequently
hired to direct a university archaeological museum.
Another serious problem is that many academics (and thus their stu-
dents) unhesitatingly cite as reality the certification of curator/archaeolo-
gists that purchased or donated artifacts are genuine and archaeological
evidence for cultural history. They also declare as archaeological fact a spe-
cific geographical provenience, a locus, for the purchased object, which
information was furnished by a dealer (although never mentioned). Mu-
seum-ritual demands that one eliminate the crucial empirical distinctions
between excavated and plundered artifacts; wbich behavior falsifies the his-
torical record. Museum curator/archaeologists also are bound by the rules
of confidentiality, rules that control all information about acquisitions.34
Museum confidentiality loyalties take precedence over archaeological loy-
alties, for which the operating word is omert.
Years ago an ex-museum curator, archaeologist and Director asked me
to look at a number of photographs of antiquities that a collector wanted
to donate to his museum: were they genuine or forgeries? My colleague
argued that if they were forgeries it would be bad for the museum, the public
and scholars to exhibit them. I agreed. To me all the objects were forgeries.
He told me he would inform his curator/archaeologist successor, and ask
him not to accept the gifts. Sometime later he informed me that his advice
was ignored. The gift was accepted because the rich collector could not be
offended. Decades ago I was asked by Teddy Kolleck, Mayor of Jerusalem,
to vet a donation of antiquities to the Israel Museum by a rich collector. I
32 See also Winter 1992: 3435; Watson 2006: xviiixix, 102, 331.
33 viz. Muscarella 2000: 2425.
34 See Muscarella 2008a: 613.
reported that all but one object were forgeries. Against the wishes of the
museums curator Kolleck correctly rejected the gift.
The examples cited here depict but one component of the Great Divide
crippling the discipline, and I turn to another group of archaeologists who
operate comfortably and successfully within the Plunder Culture. Function-
ing for decades, they operate successfully as a Fifth Column within the
archaeological realm, essentially covertly and unrecognized by many col-
leagues and the public. Most scholars aware of their roles are either indiffer-
ent or remain silentthe latter because they fear their powerful academic
connections in the awarding or denying academic positions and fellowships.
I give some examples:
University employed archaeologists accept invitations from the Metro-
politan Museums Department of Ancient Near Eastern Art to become mem-
bers of their Visiting Committee, which meets several times a year for
socials, lunch, private discussions, lectures, and private viewings of antiq-
uity purchases and gifts. The Committee is composed of museum curators,
dealers and antiquities collectors, and a good number of archaeologists.35
They are invited to become members by the museums Trustees, after which
they receive a letter from the museums legal Counsel announcing that the
Museums Board of Trustees has appointed them members. One promi-
nent archaeologist wrote in his acceptance letter: I certainly appreciate this
connection with the Museum [and] I am looking forward to the contin-
uing association [and] hope to be of some help in a visiting, admiring
and advisory capacity (italics mine). Another wrote that membership was
rewarding both on a personal and professional level (italics mine):
the precise nature of this professional engagement was manifested when in
a letter to an archaeologist colleague he referred to his official connections
with the museum that obliged him not to discuss an archaeological matter.
Other archaeologists wrote the more typical I accept with pleasure ( to
sit alongside my fellow-members, powerful and rich antiquity dealers and
collectors).
Noted for years in the invitation letters is that the Committee Chair is
Shelby White, possessor of hundreds of plundered antiquities. Other mem-
bers include the antiquity collectors Jonathan Rosen (also an antiquities
dealer), Sheldon and Barbara Breitbart, and Michael Steinhardt. In Forbes
Internet report (12/29/06: 4) the latter, defending his purchasing activities,
35 See Muscarella 2008a: 613, 616617; for a full listing of members see the Museums
Annual Reports.
boasted that he is not intimidated (what billionaire is?), that collecting was
dangerous, but that makes it exciting.
The archaeologist members serve their colleagues loyally: omert again.
They provide their dealer and collector colleagues with lectures on their
excavations and research. They visit their homes for social gatherings, and
offer advice on their purchases. An archaeologist member was also asked
by the museums Counsel to contact colleagues and officials of Turkey as
the Museums representative (another official function) to present their
position in a matter that concerned a large group of plundered artifacts, the
so-called Lydian Treasure, which the museum had purchased and refused
to return.
A few years ago I learned that the President of the NYC chapter of the
Archaeological Institute of America had voluntarily accepted the Trustees
invitation to become a Committee member. I informed her of its dealer and
collector membership, supplying their names, and asked why she would
associate with them. This archaeologist shouted that the archaeologist X
was also on the Committee, and stalked away. Years later she piously wrote
in the Fall 2007 AIA New York Society News: The AIA has worked hard to
stop the trafficking of antiquities and established itself at the forefront of
preserving the worlds archaeological resources and cultural heritage. No
footnote disassociated her from this claim (after all, she knew that a past
AIA President has been a member of the Visiting Committee for more than
three decades).
Over many years non-member archaeologists have been invited by the
Committee to give lectures on their fieldwork. With ease these archaeolo-
gists accept the honor to address White, Steinhardt, Rosen, etc. Once when
Joan Oates accepted an invitation to lecture about her site Nimrud before
the Committee, I was told by the head of the ANE Department not to come (I
had no intention to do so): because Shelby White would leave if I appeared,
and that could not be tolerated. Another time when Ian Hodder was speak-
ing at a ticketed meeting at another institution for which the Committee had
reserved 25 places, the same authority informed me (through an intermedi-
ary) that I could not attend; no ticket was available for me. In both cases,
an archaeologist was denied permission to hear an archaeologist speak, but
the dealers and collectors were not denied: on orders from the plunderers.
Why did these archaeologists consider it appropriate to give archaeological
lectures to plunderers?
Collaborationist scholars also openly support collectors and facilitate
their plunder entering the United States. A cuneiformist from Cornell Uni-
versity helped Jonathan Rosen import recently plundered tablets from Iraq
(in 2000) by informing Customs that they were educational material with
no commercial value.36 Rosen took a $1,000,000 tax deduction when he sub-
sequently gave the tablets to Cornell; and the collaborating cuneiformist will
publish them.
Archaeologists also support plundering and the selling of the orphaned
antiquities to dealers, museums and collectors, by opposing all legislation
attempting to stop them. This was evidenced in Washington D.C. during
the congressional hearings on antiquity legislation in the 1970s, where over
some days I encountered archaeologists who vigorously opposed any anti-
plunder testimony before Congress. In Israel the Antiquities Traders Associ-
ation campaign against stopping plunder and the existence of dealers shops
was joined by a number of prominent local archaeologists, and (needless to
add) museum curators.37
Others openly serving the plunderers are the many archaeologists who
write articles for the plunder/dealer-supporting magazine owned by an
antiquities dealer (Minerva), or those functioning as fronts for antiquity
dealers and collectors (Biblical Archaeological Review, Archaeological Odys-
sey). Known to these contributing archaeologists is that the dealer-owned
magazine, aside from adversary behavior toward archaeologists, in each
issue displays many advertisements from antiquities dealers all over the
world, and informs its readers about antiquity auction sales worldwide. The
front-magazines (one has ceased publication) contain advertisements from
antiquities dealers (reduced in recent years), consistently defend dealer
and collecting activities and attack anti-plunder archaeologists and their
professional organizations. Not one issue of the three magazines is without
articles by archaeologists, who report (for a fee) on their excavations or other
research matters.38 Archaeologists also serve on their Editorial Boards.39
36 Internet, Michael van Rijn, March 23, 2003.

37 Ilan et al. 1989: 40.
38 A report written by an archaeologist member of the McDonald Institute for Archaeo-
logical Research at Cambridge appeared in a recent publication in Minerva 2006; (Muscarella

2008a: 610611). Further the new Director of The University of Pennsylvania Museum of
Archaeology and Anthropology, Richard Hodges, was appointed (in 2007) to this position by
the President of the University (guided by an alleged Search Committee) to govern this said-
to-be archaeological museum: despite the Committees prior awareness that he had written
many articles and site reports over the years for Minerva, and moreover, is a member of its
Editorial Advisory Board (see below). It gets worse: one of the Search Committee members is
an archaeologist and the President of the Archaeological Institute of America (on which see
also above).
39 See footnote 38.
Some archaeological museums make public announcements of their

alleged anti-plundering attitudes, viz. the University of Pennsylvania
Museum in 1970, whose staff voted a (mild) Philadelphia Declaration
against purchasing looted artifacts. But from day one the declaration was
ignored (sabotaged) by several members of its own archaeological staff
one a former director (Robert H. Dyson, jr.): museum staff collaborated with
dealers and collectors then, and continue up to the present.40
An advertisement for a Festschrift to be published in honor of a major
serial collector (Moussaieff) has articles by archaeologists and other schol-
ars of antiquity, voluntarily honoring the collector. Archaeologists, cunei-
formists, numismatists, and art historians also authenticate and provide
cultural attributions for plundered antiquities owned by collectors, muse-
ums, and dealers; Coggins41 was one of the first to call them accomplices.42
They do this for the social prestige derived from being consulted by the very
wealthy and influential, or for grants they seek or receive from them for their
excavations.43
Archaeologists also propagandize for collectors. A university professor of
archaeology has for years positively supported antiquity dealer and collec-
tor activities, which includes securing professional positions for his graduate
students from both groups. Another, whose excavation budget has been
financed for years by a fund established by major plunderers, never dis-
cusses his sponsors antithetical-to-archaeology activities; but he publicly
defends them. Other archaeologists who raise money from a plundering
museum equally accept the requirementto support the museums pur-
chasing activities (see above). One might legitimately argue that it is not
wrong or immoral for an archaeologist to accept money from plundering
sponsorsprovided that one not avoid rejecting their activities. An exam-
ple is the archaeologist K.D. Vitelli, who received money from the Levy-
White publication fund, but did not stop her attacks on their plundering
activities. For this non-collaborationist, non fifth-columnist behavior, Her-
shel Shanks mocked her in his parti-pris magazine (funded by L. Levy and
S. White) Biblical Archaeological Review (March/April 2000: 15). Not one
40 See Muscarella 2008a: 612613; and footnote 38 above.

41 Coggins 1972: 264265.
42 Muscarella 2000: 4, 5, 12, 13, 28, footnote 11.
43 Coggins 1972: 265. I myself have not authenticated an artifact for a dealer or collector,
or participated in the purchase of one post 1970/1 (pace the self-serving and false claims by
an antiquities dealer), the year when I first learned from the lectures of Clemency Coggins
and Ezat Negahban what purchasing antiquities meant; and when I was instantly converted.
of this magazines archaeologist-editorial board members or authors ob-

jectedlet alone resigned. I know another scholar who accepted financing
from this same fund, but has not ceased to attack theirand that of other
collectors and dealersactivities.
A model of the archaeologist/curator dichotomy (pace B. Rose, above),
its Great Divide, is manifested by a meeting of archaeologists in Baghdad in
1994 where twenty-three archaeologists signed a petition condemning the
plunder and destruction of sites in Iraq. Months later one of the signers, an
archaeologist and museum curator, wrote to a curator at another museum
asking that she ignore his signature on the petition. It was only a declara-
tion of intent, and no one would expect museum employees to be bound
by such restrictions. He proudly noted that the Trustees of his museum gov-
ern him; it is they who formulate policies on these matters, they who must
be obeyed.44 Written with ease, one museum curator to another, he openly
and readily discusses deceit and betrayal, both to archaeology and the peti-
tion co-signers. The letters importance was underscored when it was for-
warded favorably to the administration of the recipients museum, which
subsequently rewarded him. Sometime later, when he applied there for the
position of Curator-in-Charge, he was appointed (although he subsequently
withdrew). Up to the present he publicly proclaims his concern for stop-
ping plunder and destruction. There exists corroboration of the appoint-
ment, and a neat explanation why this archaeologist-curator was offered it.
Another archaeologist reported to several colleagues via the Internet that he
too had applied for the same position, but was rejected: the museums direc-
tor informed him that I might be too scrupulous about acquisitions , a
concern patently not in place for the one who was offered the position.
Relevant, but little known, is that a number of United States university
museums actively purchase plundered antiquities, i.e. support the plun-
dering and destruction of ancient sitesand a number of their curators
are archaeologists. For examples of Harvards museums plunder-sponsoring
policy, see BrodieRenfrew 2005: 353, and for its internal codewhich has
been ignored (Winter 1992 does not discuss this), see Coggins 1972: 266;
Graepler 2004: 118; Muscarella 2008a: 612. Note also the ignored Philadelphia
Declaration (above). C. Nagin (1986: 23) quotes the collector Arthur Sack-
ler, who with his purchased antiquities furnished a newly created Harvard
museum that bears his name: this is a temple [sic!] of art and the art
44 Muscarella 2000: 25.

historians are its high priests, and in which temples plundered artifacts
become art. It is not a coincidence indeed that one of the owners of Har-
vard University, James R. Houghton (serving on its Board of Overseers) is also
one of the owners of the Metropolitan Museum of Art (as member and chair
of its Board of Trustees). For years in both institutions he has encouraged
the purchase of art.45 In the early 1900s Yale Universitys museum bor-
rowed many artifacts from Peru, from Yales excavations at Machu Picchu.
Yale never returned them to Peru and as of 2008 continues to refuse repatri-
ation.46
Other plunder-active United States university museums include Prince-
ton, Indiana University, University of Missouri, Emory, etc 47 University
trustees, presidents, faculties, including archaeologists, support these pur-
chases, actively or by their silence: they do not write letters to their univer-
sity presidents demanding an end to their participation in the destruction
of the archaeological record. For European University museum participa-
tion in the Plunder Culture, see Graepler 2004.
Conclusions: The archaeological Fifth Column is a significant component
of the forces that sponsor and thrive on cultural destruction. Its members
are as equally culpable as the advancing columns, tombaroli, antiquity deal-
ers, collectors and museum staffs. They continue to flourish, their activities
proceed successfully and unabated, they get awardedrevealing that the
discipline of archeology has no comprehensive sense of itself, no unclouded
self-knowledge, no awareness of its moral and academic weakness.48
There remains a parallel conclusion: follow the archaeological work, the
integrity and accomplishments of archaeologists like Altan Bey, whose
actions manifest proper archaeological behavior.
ok tesekkr ederim, arkadasm.
45 In a classic conflict of interest action Houghton wearing his Harvard hat awarded an
honorary Ph.D. to his Metropolitan Museum employee de Montebello, a degree this director
could/would never have earned.
46 E. Karp-Toledo, The Lost Treasure of Machu Picchu, New York Times, 2/23/08: A17.
47 Muscarella 2008a: 612614; Winters (1992: 3536) comment that university museums
are reluctant to purchase plundered artifacts is incorrect.

48 Every event and issue cited in this paper derives from my own research and experi-
enceshence the often usage of the pronoun I. There are surely many other examples of
which I am unaware.
Bibliography
Atwood 2007: R. Atwood, A critical look at U.S. media coverage of antiquities issues,
SAFE (Internet), 17.
BrodieRenfrew 2005: N. Brodie, C. Renfrew, Looting and the Worlds Archaeolog-
ical Heritage: The Inadequate Response, Annual Review of Anthropology, 343
428.
Brodie et al. 2000: N. Brodie, J. Doole, P. Watson, Stealing History: The Illicit Trade in
Cultural Material, Cambridge.
Brodie et al. 2001: N. Brodie, J. Doole, C. Renfrew, Trade in Illicit Antiquities: The
Destruction of the Worlds Archaeological Heritage, Cambridge.
Brodie et al. 2006: N. Brodie, M. Kersel, C. Luke, K.W. Tubb, Archaeology, Cultural
Heritage and the Antiquities Trade, University Press of Florida.
Coggins 1972: C. Coggins, Archaeology and the Art Market, Science 175, 263266.
Graepler 2004: D. Graepler, Archologie und illegaler Antikenhandel: Die rolle der
Universittssammlungen, W. Heilmeyer, J.C. Eule (Eds), Illegale Archologie?,
Berlin, 116127.
Griffin 1986: G.G. Griffin, In defense of Collectors, National Geographic 169, 4, 462
465.
Griffin 1989: Collecting Pre-Columbian Art, in Messenger 1989, 103115.
Horsley 1997: C.B. Horsley, Chinagate, The City Review: 116.
Ilan et al. 1989: D. Ilan, U. Dhari, G. Avni, Plundered!, Biblical Archaeological Review
XV, 2, 3842.
Mackenzie 1988: S.R.M. Mackenzie, Going, Going, Gone: Regulating the Market in
Illicit Antiquities, Institute of Art and Law, Leister.
Mazur 2006: S. Mazur, Italy-Met Euphronios Accord? Scoop Internet, February 22:
112.
Meyer 1974: K. Meyer, Behind the Damask Curtain, MORE: 1114.
Messenger 1989: P.M. Messenger, The Ethics of Collecting Cultural Property: Whose
Culture? Whose Property?, Albuquerque.
Muscarella 1977: O.W. Muscarella, Ziwiye and Ziwiye: The Forgery of a Prove-
nience, Journal of Field Archaeology 4, 197219.
Muscarella 1991: O.W. Muscarella, Review of Messenger 1989, American Journal of
Archaeology 95, 2, 342344.
Muscarella 2000: O.W. Muscarella, The Lie Became Great, STYX.
Muscarella 2003: O.W. Muscarella, Museum Constructions of the Oxus Treasure:
Forgeries of Provenience and Ancient Culture, Ancient Civilizations from Scythia
to Siberia, 8, 34, 259275.
Muscarella 2006: O.W. Muscarella, Urartian Metal Artifacts: An Archaeological
Review, Ancient Civilizations from Scythia to Siberia 12, 12, 147177.
Muscarella 2008a: O.W. Muscarella, Archaeology and the Plunder Culture, Inter-
national Journal of the Classical Tradition, 14, 3/4, 602618.
Muscarella 2008b: O.W. Muscarella, The Veracity of Scientific Testing by Conser-
vators, Original, Copy Fake?, Mainz, 918.
Nagin 1986: C. Nagin, Patrons of Plunder, Boston Review, 525.
RoseBurke 2004: J.D. Rose, D. Burke, Making Money from buried treasure, Culture
Without Context, 14, 48.
Rose 2007: C.B. Rose, From the President: Common Ground, Can Archaeologists
and Museums Work Together?, Archaeology 60, 5.
Simpson 2005: E. Simpson, Tall Tales: Celts, Connoisseurs, and the Fabrication of
Archaeological Context, SOURCE XXIV, 2, 2841.
Watson 1977: P. Watson, Sothebys: The Inside Story, London.
Watson 2006: P. Watson, The Medici Conspiracy, New York.
Winter 1992: I. Winter, Changes in the American Art Museum: The (An) Art His-
torians Voice, in Different voices: a Social, Cultural, and Historical Framework for
Change in the American Art Museum, Association of Art Museum Directors, 30
57.
Section Four
Forgeries
chapter thirty-two
BAZAAR ARCHAEOLOGY*
(PLATE 34 A.B)
It is unfortunately not erroneous to state that a high percentage of the infor-

mation, conclusions, reconstructions and ideas published about the art and
archaeology of the ancient world is (one or more of the following) utterly
wrong, contrary to common sense, based on uninvestigated claims or false
data, and conducted without regard for proper archaeological methodology.
For the latter one thinks of asking how, when, where, why and what of all
the material available for study, questions that should be asked all the more
forcefully of non-excavated material under investigation. But these ques-
tions rarely get asked and the anything goes and anything-is-published
syndrome is standard behavior in the discipline. It is difficult to avoid pes-
simism and a cynical attitude toward current approaches to the study of
ancient art. As no end to the problem is in sight, why bother to report on
yet more solecisms presented as knowledge of the ancient world? But then
one thinks of scholars like R.M. Boehmer who has given decades of his life
to the task of cautiously interpreting archaeological evidence, and who with
thoroughness, intelligence and a hard-earned knowledge of facts, attempts
to reveal as accurate a history of the past as is objectively possible. Knowing
that scholars like Boehmer are at work in the field makes it possible to con-
tinue to believe that perhaps the discipline of archaeology is worth fighting
for.
In this context I discuss two small lapis lazuli discs, carved on both sides
in apparent Early Dynastic style, that surfaced in the bazaar (the antiqui-
ties market) nearly two decades ago. Both have generated positive atten-
tion in print and scholarly dialogue. Because the discs are clearly related
in form and style, and surely share the same purpose that determined
their manufacture, the issues and conclusions advanced concerning their
alleged ancient histories deserve close examination. Equally deserving of
close examination is the methodology inherent in these discussions.
* This chapter originally appeared as Bazaar Archaeology, in Beitrge zur Kulturge-
schichte Vorderasiens, Festschrift fr Rainer Michael Boehmer, eds. U. Finkbeiner, R. Dittmann,

and H. Hauptmann (Mainz: von Zabern, 1995), 449453.
880 chapter thirty-two
The first disc (pl. 34 a; diameter 2.6cm., thickness .8 cm.) surfaced in 1976
and has been published in two dealers sales catalogues and subsequently
in two scholarly venues; it now rests in a New York private collection.1 In its
first non-dealer publication (see note 1: Porada 1992; the second publication,
Porada 1993: 50ff., is less detailed), the iconography and style of the carved
scenes are concluded to date to the Early Dynastic period. On one side is
depicted a seated male figure wearing a tufted skirt with his right hand clasp-
ing the left; before him is a cuneiform inscription that several cuneiformists
read as Rimush LUGAL KISH, that is, the Akkadian king who ruled about
ca. 2270 bc. The other side depicts an Anzu-Imdugud bird monster grasping
in each claw a horned animal. The rim is pierced for suspension.
Concerning the seated figure it is noted (Porada 1992: 69), with Early
Dynastic style in mind, that the head is very large, in proportion to the
body, the eye is huge, and, more unusual the ear is very large, as is also
the nose; and only one foot is indicated With regard to the bird monster
1 Collectors Journal of Ancient Art, Vol. 5, no. 2, 1985, front and rear cover; Hesperia
Arts Auction Ltd. Antiquities, November 27, 1990, no. 58; E. Porada, A Lapis Lazuli Disc with
Relief Carving Inscribed for King Rimush, in La circulation des Biens, des Personnes et des
ides dans le proche-Orient ancien: Acres de la XXXVIIIe Rencontre Internationale (1992)
69 ff.; idem, Seals and Related Objects from Early Mesopotamia and Iran, in J. Curtis (Ed.),
Early Mesopotamia and Iran (1993) 44 ff.; neither article mentions the other. The disc is also
listed in D. Frayne, Sargonic and Gutian Periods (22342113bc), RIM volume 2 (1993) 71, no. 51
(incorrectly said to be in the collection of the P. Morgan Libraryit is only on loan there).
Note that no. 52 in the RIM volume is the second disc discussed by me in this article, infra.
I also wish to thank J. Rosen for allowing me to examine the disc, and David Wright of the
Morgan Library for facilitating the examination.
According to Porada (1992: 69) the disc came to its present provenance from the dealer
(the former owner) Joel Malter, who got it from another dealer who had acquired it from an
old European collection. Aside from the problem of the last undocumented claim, which in
most casesas every plunderer, smuggler, dealer, collector, curator, knowsis either gossip
or outright dissimulation, usually the latter, the given trade sequence is incorrect and much
telescoped. Indeed, the actual modem trickle-down trade dynamics is more complex. In 1976
a dealer, X, purchased it from another dealer. X then re-sold it to another dealer, who claimed
to be acting as agent for a client. At sometime after X had purchased and sold the disc, he was
told that it was once in the Gillet collection, but no evidence for this claim, or information
about how long the disc allegedly existed in the collection, was given to him. Nine years
after X sold it, in 1985, the disc was put up for sale (Malter), with no information given about
previous owners. Five years later the disc surfaced again, in an auction catalogue (Hesperia:
which is not mentioned by Porada); it then went to its present home in the J. Rosen collection.
Thus, before its final sale at least five, six, possibly seven (if we count the old family
collection) transactions, and possibly more not recorded (between 1976 and 1985, and 1985
and 1990?), had occurred; and one assumes that the price had risen as many times. Do we
have here an ethnoarchaeological model for ancient trade?
I also wish to thank Elizabeth Simpson for reading this paper, for making good sugges-
tions. and for demanding that I remove many preaching sentences.
bazaar archaeology 881
scene, it is noted (ibid.: 70) that the grasped goats with long, plain horns
are not paralleled in Mesopotamian cylinders and that the depicted
monster-bird differs from all Mesopotamian representations by having
the head of a bearded man. Also noted is that the dot circles on the rim
were made with a mechanical tool. Following these correctly observed
deviations from that which is known to be Early Dynastic (which is the
existence of a fairly large amount of excavated material),2 formal parallels
that allow the disc to be dated to a post-Urnanshe period are drawn (Porada
1992: 70; 1993: 50). The date suggested for the disc is the period of either
Eannatum I or II, ca. 2450bc, despite the authors opinion that the presence
of one row of tufts at the base of the gown is an indication of a pre-Urnanshe
date (ca. 2470bc; 1992: 70).
In fact, this list of deviations from that which is known to be Early
Dynastic (see note 2) does not reveal the full extent of the problem. The
observant viewer will readily add the following (some are more visible in
autopsy than in photographs): not only are the nose, ears, and eyes too large,
they are crudely executed, and in forms unparalleled and unexpected in the
extant corpus: note the eye outline that curves back and then loops forward
to incompletely enclose the pupil, and also the awkward juxtaposition of
the eye and ear. The large head lies heavily on the shoulders, and the curved
mouth in the projecting chin is clamped, not merely closed. The body, single
foot (a blob) and arms, and the bottom row of skirt tufts are not executed
in typical Early Dynastic forms. The hands have no structure, the fingers
are not discernible, and instead of a thumb extending from each of the two
hands, a thumb (?) and second finger awkwardly project from the under, left
hand. (Actually, all one need do is compare this figure to Porada 1992, Fig. 2
to appreciate the significant differences.) Also, the figure sits not on a stool
with a curved top but unsurely inside a chair. The dotted circles along the
exterior rim of the disc are unevenly spaced and some overlap.
And there is more. It is unusual for a seated figure represented on Early
Dynastic reliefs to be depicted with hands clasped; this gesture seems to
be reserved for standing figures in relief and in the round; and the holes
indicating the nipples do not to my knowledge occur elsewhere on relief
2 For purposes of examining the references for the appropriate excavated Early Dynastic
material, see for example H. Frankfort, Sculpture of the Third Millennium B.C from Tell Asmar
and Khafajah, (Chicago, 1939): pls. 105114; idem, More Sculpture from the Diyula Region
(Chicago, 43): pls. 6367; A. Moortgat, The Art of Ancient Mesopotamia (London, NY, 1969):
pls. 35, 4249, 109121; I. Fuhr-Jaeppelt, Materialien zur Ikonographie des Lwenadlers Anzu-
Imdugud (Mnchen, 1972), passim.
figures, rather they seem to be confined to statues in the round to hold inlays.
As for the deviations mentioned above for the bird monster, aside from the
presence of the goats and the beard (it is definitely undercut to indicate a
beard and is a misunderstanding of the hatching directly below the felines
nose-mouth area noted on most representations of the creature), note the
crudely drawn wings, the truncated bodies of the grasped animals that sink
below the edge, and their indistinct heads.
The absences of parallels in style, iconography, and execution are all the
more telling when one considers the essentialbut never mentioned
fact that no lapis lazuli disc with an Early Dynastic scene has ever been
excavated, anywhere.
Even keeping in mind only the short list of deviations mentioned in
the two articles, one wonders why it did not trigger questions or doubts,
especially considering the objects lack of an archaeological history. Several
answers may be suggested for the omission, but one is explicit in the pub-
lications: it was uncritically accepted as a given, as an inherent truth, and
hence assserted, that the dealer-derived, unexcavated disc originated (i.e.
was manufactured) in the Kerman area of southeastern Iran (Porada 1992:
69f., 72; 1993: 50ff.). The evidence offered for the assertion that an object
with an Early Dynastic design was manufactured outside of its expected
homeland was that the disc resembles another, also unexcavated, lapis lazuli
disc (Louvre, ex-Foroughi Collection; Porada 1993: Fig. 19). This example
depicts what seems to be a scene more at home in Iran than elsewhere, and,
following P. Amiet, is assigned to the Kerman area. In this mode of analy-
sis, the provenience of one unexcavated disc is taken to be the same as
the provenience of another unexcavated disc, differences in style notwith-
standing. Further, the deviations in style and execution from the corpus
of Early Dynastic artifacts have no authority other than as obvious (after
the fact) evidence and support for a perceived non-homeland source. Thus,
even before an Iranian origin was proclaimed (Porada 1992: 70), it was
stated without evidence and matter-of-factly in defense of one deviation
that it is probably a type at home in southeastern Iran. And the histori-
cally significant conclusion that the disc was made in a non-Mesopotamian,
provincial area was posited before stylistic discussion commenced, provok-
ing the recognition of a close relation to the Mesopotamian works and
their unquestioned combination with foreign elements , the latter being
the deviations noted (ibid.: 72). From the beginning, the inscription was
accepted without question as a genuine component of the ancient disc.
The ethnic and geographical backgrounds of the disc having been estab-
lished, a suitable historical scenario was then devised to support them: the
disc was created with reference to Early Dynastic Mesopotamian art by a

contemporary, local sculptor in Kerman about 2450 bc. The disc became an
heirloom, indicated by the inscription added by the Akkadian king Rimush,
son of Sargon, who ruled some 200 years after the time of the discs manufac-
ture. Since Rimush is the very king who defeated Marhashi, a polity accepted
without reservations to encompass the Kerman area, and where, remember,
it has been determined the disc was made, it naturally follows that the disc
must have been booty taken by the Akkadian King from Kerman.3 Rimush
doubtless had [it] inscribed when he obtained it in about 2270bc and
then presumably brought it back to Mesopotamia (1992: 72; 1993: 52). Thus,
the phenomenon of the disc itself is sufficient archaeological evidence to
determine its place of manufacture and ancient travels.
Put another way, a craftsman in Kerman is commissioned for unknown
reasons to carve Early Dynastic-style figures on a lapis lazuli disc; he does
his best, but being a provincial and not half so good as a Sumerian, he
misunderstands everything; nevertheless, local tastes determine that the
disc is valuable and it is preserved; 200 yearsor five to seven generations
after its manufacture an Akkadian king at great risk and travail crosses
extraordinarily difficult terrain, reaches Kerman, finds the treasured disc,
considers it worthy both as booty and as meriting his signature, takes it back
home to Mesopotamia, where it remains in the ground until 4200 years later
plunderers find it and smuggle it by long distance trade to the United States.
There is, however, one more piece of relevant information that has hith-
erto not been noticed: the inscription is problematic. Two cuneiformists
who examined the inscription observed that it would/could not have been
written by an ancient scribe.4 Douglas Frayne wrote: The first sign in the
3 Opinions concerning the whereabouts of Marhashi are still being expressed, and schol-
ars disagree about its geographical position: see for example the article by P.R.S. Moorey in
the Curtis 1993 volume mentioned in note 1, p. 38; wherever it existed, it has never been
demonstrated that any Mesopotamian army ever penetrated into Iran so far as, or even close
to, Kerman. In both sales catalogues (see note 1) it is stated (with scholarly help?) that the
disc was manufactured in 2800bc and that Rimush may have or probably acquired the
disc as booty when he conquered Kish more than 400 years later. Speaking only to these
problems, first, there is no evidence that Rimush conquered KishSargon probably accom-
plished that feat. And, second, the translation of the inscription is probably: Rimush, King of
the World/King of the Universe/Knig des Alls (as Frayne and others have noted).
4 Three cuneiformists examined the inscription at my request; because if the disc could
be a forgery then perhaps the inscriptions might reveal some mistake. Two of the cuneiform-
ists had previously seen it and thought it genuine (one was D. Frayne, see note 1), but
after reexamination they changed their minds. The thirda tenured professor in a major
universityrefused information because he said he did not want to offend the discs owner
lapis-lazuli disk, as far as can be determined from the photo published

seems to be an URUDU or possibly a MES sign rather than the expected
URU sign. One cuneiformist maintained that the signs are crudely executed
and probably are not ancient. Although it surely would have been a (royal)
scribe (not an illiterate artisan) assigned to write Rimushs inscription on the
disc, one cannot resist the temptation to ask whether it could be argued that
he must have been either a beginning scribe (l.hu.ru.(um)), or an unskilled
scribe (tupsarrum hurrum)or perhaps a provincial scribe from Kerman?

The second lapis lazuli disc depicting an Early Dynastic-like design is
in the Bible Lands Museum, Jerusalem; it has been possible to trace its
first appearance to about 1979/1980 when it was purchased by its present
owner. To date, it has not been published, but it has been seen (autopsy or
photographs) by a number of scholars (pl. 34 b)5 and discussed; in size it is
a little larger than the first example. On one side are two bare-chested male
figures. One, wearing a full fleeced skirt, is seated and facing right; the other,
wearing a plain skirt tufted only at the bottom, stands before him. Each
holds in one hand (the left of the seated, the right of the standing figure) the
feet or tail area of an Anzu-Imdugud bird monster, which forms the scenes
center and its focus of attention. A cuneiform inscription is incised between
the figures and on the skirt of the standing figure; it reads Rimush LUGAL
KISH. On the other side is an Anzu-Imdugud bird monster grasping two
felines with his claws.
The seated figures head joins his shoulder without a neck; he has a
hooked nose; his right hand is a lump that holds a curved object that has
no delineated terminal; a round object (meant to be a knife handle?) seems
to project vertically from his wrist; the tufts on his skirt are irregular and
slope down; the feet are short, squared lumps; the stool is cut off at the back
and seems to slope from bottom to top; and the stool rests on a thin strip
that itself rests on a cross-hatched platform. The standing figure also lacks a
neck; the facial features are barely articulated, but they are not those of his
colleague: the nose is larger; the left hand seemingly holds or touches some
long vertical, amorphous raised area at the chest (is there restoration here?);
and the one foot preserved is also stumpy.
and its publisher (nevertheless, he inadvertently answered my question). This situation is of

sociological interest for what it says about the intellectual independence of the discipline.
Note also that I have been informed that Erica Reiner never saw the inscription (v. Porada
1992: 72).
5 It is a puzzle why Porada, who knew of the Jerusalem discshe was shown it in about
1980did not mention it in either of her two articles.

Each feather of the bird monster on the other side is delineated by crude
oblique lines; the wings are truncated and unevenly curved on each side; the
ears are pointed and cat-like. The felines held in the birds claws are depicted
like those on the Dudu plaque from Tello, with the full face seen from the top
facing the bird, but here not so satisfactorily accomplished: on the disc the
heads are not a pair and are crudely executed; the same is observable for
the asymmetrical front and rear legs and truncated feet; and the tails hang
straight down.
Disregarding that many of the discs most striking features, details, and
iconography, are unparalleled in Early Dynastic (or any other ancient
periods) art, an interpretation has been vigorously offered (in communica-
tions and dialogues) for the iconography is that it is a major example of polit-
ical propaganda created in the Akkadian period (argued thusly because of
the inscription): a bearded Rimush stands equal in height to and facing the
unbearded deity Ningirsu (on the unproven assumption that the Anzu bird
is the symbol primarily of this deity). The grasping of the Anzu-Imdugud
bird monster by both of the figures represents their sharing of earthly and
divine power, truly an extraordinary political event and an unprecedented
scene in ancient Near Eastern art. The execution of an Akkadian period
object in what is normally called Early Dynastic style is perceived in this
interpretation as an overlap of the new political-cultural authority with the
older, but still present, styleand the disc is cited as the very evidence
demonstrating this phenomenon. Questions raised by concern for the lack
of parallels for style and execution of the unique iconography produce the
response that no forger could have conceived the ideologically exquisite
message of Machtkunstwhich argument automatically absolves the stylis-
tic and chronological issues from consideration. Indeed, the argument is
precisely that the lack of stylistic and iconographical parallelsin both the
Early Dynastic and Akkadian periodsis the very evidence that indicates
the discs uniqueness and therefore its singular importance as an ancient
object.
If one accepts the disc as an Akkadian product, one must confront both
the presence of the apparent Early Dynastic style and the absence of any
other evidence of the continuation of that style into the Rimush period (not
to mention the earlier ca. 60-year reign of Sargon). An Akkadian attribution
can be maintained only if the inscription can be demonstrated to be both
ancient and contemporary with the depictionthat is, not to have been
added two centuries after the carving. On the other hand, if it is assumed
that the disc was made in the Early Dynastic period, then both the aberrant,
unparalleled, style and the unique iconography and ideology, although not
the inscription, would have to be accounted for, just as with the first disc
discussed. One thing is certain, however: the proponents of either view will
need to invoke the ubiquitous Deus ex machina known as the Provincial
Craftsman.
The history of both lapis lazuli discs begins in the bazaar, and all evidence
indicates that their style and iconography are unparalleled. Why then, and
by what rules of archaeology/art history, are we required to accept them
as genuine, ancient artifacts? Can they be demonstrated to be ancient?
Clearly no. And it is not so much that their antiquity has decidedly not
been demonstratedlet alone provenbut that no verifiable evidence for
their antiquity has been presented. Given the lack of real evidence for their
antiquityin this case evidence that would have been provided had the
discs been excavated, the possibility, indeed probability, that the objects are
not ancient must be considered.6
Correctly understood, the dilemma of the lack of verification takes us
beyond specific arguments that the discs may be or are forgeries. For an
explicit consequence of their condition is that although some scholars may
honestly argue that the discs are genuine, or that they are not sure the discs
are forgeries, the fact remains that the discs have ceased to be, and cannot
be cited as, artifacts known to have been made in an ancient culture. For,
even if it is argued that it cannot be proven that the discs are forgeries, it
equally cannot be proven that they are genuine. This is significant, and it
manifests that the discs are lost to archaeologyeven if in fact they are
genuine! Such is the fate of all similarly situated objects that surface in the
bazaar and exhibit unique, anomalous characteristics that are unparalleled
6 Taking into consideration the possibility that the history of the discs exists solely in the
modern world, some inferences emerge. The discs were probably made in the early 1970s
in the same workshop (in Iran?) but apparently by two artisans. The forgers either knew
of decorated lapis lazuli discs (two are known: one already mentioned in the Louvre; the
other excavated in EgyptPorada 1993: figs. 18, 19) and used them as models for the new
creations, or they acquired genuine (plundered) blanks that begged for decoration (labo-
ratory analysis may clarify this issue; also needed is an analysis to verify that the material
is lapis lazuli). The forgers attempted to replicate generic Early Dynastic scenes. They bor-
rowed, adapted and enhanced features from the often published Early Dynastic iconography,
viz. decorated stone plaques and the Eannatum Vulture stele. Somehow (through cooper-
ation with a cuneiformist or by library research) the forgers came across the many exam-
ples of Rimush inscriptions and copied them. The Jerusalem discs inscriptionif not the
otherhas been vigorously defended as authentic, but this authenticity may be in a formal
sense, that is, it could have been expertly copied. The forgers may not have been aware of
the chronological/cultural discrepancy involved between Early Dynastic style and Akkadian
inscriptions: or perhaps they did and planned to be clever.
in the ancient corpus. This fate is the gift of the bazaar to modern culture,
the claim that all material offered for sale is genuine plundered artifacts,
gifts from the broken earth.
The bazaar model requires that we accept a methodology that incor-
porates unanchored intuition and a priori faith, and rejection of the facts
within archaeologys care. This methodology instructs us that stylistically
un-canonical, idiosyncratic, and poorly executed artifacts that have their
origin not in mounds and site reports but in dealers shops and catalogues,
are ancient artifacts, and that scholars have a duty to study the bazaars,
not dig in excavations, to acquire from these aberrant objects meaningful
knowledge about the history and culture of the ancient world. This is a
model that does not believe archaeological culture is worth preserving, that
it hasor deservesa future as a serious scholarly discipline.
a. The first lapis lazuli disc (New York private collection).

b. The second lapis lazuli disc (Bible Lands Museum, Jerusalem).
chapter thirty-three
EXCAVATED IN THE BAZAAR: ASHURBANIPALS BEAKER*
Bazaar archaeology flourishes. Its advocates in museums and universities

profess the discipline to students and the public. Bazaar archaeologists
have the advantage of investigating more accessible and less complex sites
than those chosen by conventional, reactionary field archaeologists. Their
unearthing occurs not in the trenches of remote, dusty sites, but in above-
ground, well-appointed bazaars (aka antiquity dealers shops) in Zurich,
Paris, London, Tokyo, New York, and so forth, and in the sumptuous venues
of museum and private collections. Herein they seek to discover the extraor-
dinary, the unique work of art. And they do.
All bazaar archaeologists share certain beliefs: that art-historical, archae-
ological, and cultural knowledge is viably obtained from data acquired
from bazaar-derived objects; that these data are equal tonay, are to be
privileged overempirical, and thus restricted, evidence obtained from
archaeological excavations; that bazaarists know more than archaeologists
do (their success seems to demonstrate the truth of this!); and that all arti-
facts that they or others unearth in the bazaar and publish are ipso facto
ancient. No wonder the discipline is so popular.1
Those who cruise the bazaars know that here are displayed innumer-
able objects said to be from ancient sites and claimed (with sophisti-
cated smiles) to be from collections formed many decades ago by respected
families. Here are found unicums in quantities far more than recognized
among site-excavated handicapped artifacts. Unicum antiquities share one
exquisite quality: they never match the styles and forms or the tectonic and
spiritual sense of excavated artifacts. Thus, they reveal to discerning schol-
ars new iconographies, motifs, and forms that yield unexpected but wel-
come information (usually that they were made by provincial craftsmen).
* This article originally appeared as Excavated in the Bazaar: Ashurbanipals Beaker.
Source 20, no. 1 (2000): 2937.

1 I first articulated bazaar archaeology in Bazaar Archaeology, in Festschrift fur R.M.
Boehmer, ed. U. Finkbeiner et al. (Mainz: 1995), pp. 449453. For a fine documentation of
its adepts defending unicums (forgeries), see K. Butcher and David W.J. Gill, The Director,
the Dealer, the Goddess, and Her Champions: The Acquisition of the Fitzwilliam Goddess,
American Journal of Archaeology 97 (1993): 383401.
890 chapter thirty-three
It is precisely these characteristics that define their aesthetic, emotional,

historicaland marketvalues and that automatically accord them an
ancient life. Bazaar archaeologists cannot contain themselves until the
antiquity is purchased and displayed and published.
The bazaar archaeologist never reflects upon or is troubled by the prove-
nience of a bazaar-generated artifact (unicum or otherwise), the fact that it
is a phenomenon surfacing in a bazaar shop: How did it get to the bazaar?
The said to come from X provenience asserted by a dealer, collector,
curatorall qualified bazaaristsis instant historical reality. But all this
notwithstanding, bazaar archaeologists have captured much of the archae-
ological realm. They have made plundering the Ersatz of archaeology. They
have seen to it that archaeology, excavation, and provenience are terms that
lose their names. As for the F word (Forgery), here, too, they dare not
speak its name.
Bazaar archaeologists may perhaps be distinguished in degrees from
scholars who do not support bazaar behavior but nevertheless publish forg-
eries as ancient artifacts.2 But both share blame, and the desire to publish the
unicum is equally supported by the plundering and forgery cultures, whose
artifacts include a long list of scholarly venues available for such publica-
tion.3
I call attention to a recent publication that neatly paradigms the basics
of bazaar-archaeology methodology and scholarship (a Bazaar Archaeology
for Dummies). In Schtze des Orients Meisterwerke aus dent Millo Museum, an
2 A discussion of the interrelated forgery and plundering cultures appears in my forth-
coming The Lie Became Great: The Forgery of Ancient Near Eastern Cultures. Many scholars,
including me, have published forgeries as genuine. In some cases, it was inadvertente.g., a
mistake in interpretationbut not necessarily resulting from innocence of bazaar behavior.
My mistakes appear in The Lie Became Great (see above).
3 Of course, excavations provide variations of known iconographies and stylese.g., a
bowl excavated at Arjan in southwestern Iran: Y. Majidzadeh, The Arjan Bowl, Iran 30 (1992):
131144 (John Curtis reminded me of this vessel decorated with unique scenes). Leaving aside
the fact of excavation, its uniqueness does not raise eyebrows, and every line, scene, etc. is
what we expect in an ancient work (here an Assyrianizing Iranian production).
See also the neo-Neo-Assyrian cylinder seal offered in Sothebys Antiquities from the
Schuster Collection (London: 10 July 1989), no. 45, depicting a siege scene; it was recently
made to be sold (like other modern seals exhibited in the same said-to-be antiquities sales
catalogue). E. Bleibtreu also published it (Pestungen auf neuassyrischen Rollsiegeln und
Siegrlabrollungen) in Beschreiben und Deuten in der Archologie des alten Orients: Festschrift
fr Ruth Mayer-Opificlus, ed. M. Dietrich and O. Loretz (Mnster: Ugarit-Verlag, 1994), p. 10,
no. 4, Abb. 5; see my The Lie Became Great, Assyrian, no. 40, correctly noting kein Beispel
bekannt keine Parallelen but not indicting it. For another neo-Neo-Assyrian seal with a
siege scene published as ancient, see J. Eisenberg, Glyptic Art of the Ancient Near East
Part II, Minerva 9, no. 4 (1998): 11, fig. 32.
excavated in the bazaar: ashurbanipals beaker 891
exhibition catalogue (proudly edited by Wilfried Seipel, Kunsthistorisches

Museum Wien, 1999), a gilt silver beaker bearing two inscriptions and dec-
orated with Assyrianlike scenes is vividly presented to the world for the first
time in an essay (pp. 20ff.) and a catalogue entry (no. 19) by Erika Bleib-
treu. Only one view is illustrated; I offer four views (Figs. 14).4 The pub-
lished drawings are imprecise, not revealing the crudeness of the execution
throughouta characteristic unmentioned.5 Serious problems surface after
a very brief examination of the decoration: one is in fact bewildered that
anyone knowing Assyrian art could accept the beaker as having been made
in ancient Assyria.
On the very first page, the author situates the dissimulation that informs
the publication by misrepresenting the modern history of the vessel. She
invokes the chance find fairy, a claim she surely knows is not true. If
ancient, it was plundered from a now destroyed site and smuggled out of its
country of origin (the origin of all genuine plundered material). But bazaar
archaeologists know that one never offends bazaarists and benefactors, and
they also know from experience that their claims will be accepted, more
than not, as historical reality (like an Oliver Stone movie).
Two royal inscriptions of different language and alleged date are in-
scribed, but no photographs are vouchsafed! On the interior rim is a Neo-
Elamite inscription of Ampirish, King of Samati, son of Dabalas; on the
outer rim (Fig. 3), that of Assurbanipal, king of Assyria (668627 bc). Schol-
ars date the Neo-Elamite inscription to between the ninth century and 550
or so bc. The Miho beaker is allegedly one unit of a large group of plundered
silver objects (230 are claimedp. 21), many inscribed with Ampirishs or a
relatives name. Examples have been offered for sale in bazaars since 1993;
others have been confiscated from the plunderers within Iran.6
4 I was generously given permission to publish these photographs by Hajime Inagaki of
the Miho Museum, who knew that I had different views regarding the vessels age. I have been
informed that a monograph on the vessel has recently been published by BleibtreuEin
Vergoldeter Silberbecher des Zeit Assurbanipals in Miho Museum: Historische Darstellung
des 7. Jahrhunderts v. Chr.-in Archiv fr Orientforschung Beiheft 28 (1999). See also Museum
Topics, Miho Museum News 4 (1999): 4.
5 An examination of small-scale figures depicted on excavated cylinder seals or on royal
garments on reliefs puts into perspective the beakers artistic crudeness.

6 For discussions of vessels with these inscriptions, see A. Caubet, Goblet: Pr-achm-
nide, Revue du Louvre 4 (1995): 81; V. Donbaz, A Median (?) Votive Inscription on a Silver
Vessel, Nouvelles Assyriologiques Brves et Utilitaires (N.A.B.U.) 2 (1996): 3739. A fanciful,
undocumented history of the find is given by the antiquity dealer H. Mahboubian, Treasure
of the Mountain (London: 1995). It resulted from an excavation conducted by an archae-
ologist showing diligencei.e., a dealer in antiquities. Others, including scholars, repeat
Bleibtreu (p. 21) explains the presence of two inscriptions: Assurbanipal

commissioned the vessels decoration and had it inscribed; subsequently, it
was acquired by the Median destroyers of Assyria in 612 bc, taken as booty
to Iran, and there inscribed by one of its new owners about 550 bc. Or it may
have been a Gastgeschenk from the Assyrian king to one of the subordinate
states in Iran.7 Thus, a bazaarists model for interpreting subtle clues: lets
write an archaeological storyits easy and fun! The intellectual frisson
is further intensified by the fact (sic) that the beaker has no parallels in
shape or decorationor double inscriptionsamong excavated Assyrian
artifacts: it is an unicum, a freiwillig gift of the bazaar.
The scenes, in four zones, depict an Assyrian king celebrating a victory
over the Elamites, whose king is represented several times. Also present are
soldiers, courtiers, captives, and musicians formally paralleled on the reliefs
of Ashurbanipal8 and of Sennacherib. But the many anomalies, inconsisten-
cies, and deviations from the realia and stylistic sense and spirit of these gen-
uine/excavated reliefsindeed, from Assyrian art in generalreveal that
the Miho scenes are a creation not of the seventh century bc, but a fiction
of the 1990sad. It is a neo-Neo-Assyrian (or, if preferred, post-Ashurbanipal,
very late Assyrian) artifact.
Consider the following. Starting at the top zones (Figs. 1, 3, and 4), note
that the musicians wear different headbands and garmentsand what are
these headbands?; some figures are bearded, others not; the double pipes
are positioned too high, and the fingering seems wrong; the depiction of
the lyre and harp forms (see Bo Lawergrens article following this), quivers,
and wild feather headdresses are unparalleled and betray modern misun-
derstanding; there are no seventh-century parallels for the checkerboard
garment decoration, also in the lower zones; the two horses being led are
the dealers report of an excavation from (we are generously given two choices) a cave in the
Zagros mountains, near Khorramabad or between Hamadan and Kermanshah (ibid., pp. 11,
18). The dealer date of discovery is pre-1934 (see the first unnumbered page and pp. 10ff.). In
Le royaume lamite de SAMATI, N.A.B.U. 1 (1996): 21f., F. Vallat concludes that the state
mentioned is Samati and that the dates are c. 585539bc.
An apparent example of a plundered cave vessel with an Elamite inscription is illustrated
in Christies sales catalogue Antiquities (New York: 4 June 1999), no. 204; it sold for $43,700.
A leaf pattern is incised in its in-curved sides. There is no longer any knowledge of the
provenience of the cave vessels.
7 Either way, the author commands us to believe that the (decorated) vessel rested in the
earth until found durch Zufallno doubt by poor ignorant peasants working in their cave
garden but canny enough to help get it to Japan.
8 Bleibtreu cites the appropriate relief literature but betrays no awareness that the depic-
tions differ essentially in all details, including iconography, from the scenes on the Miho
beaker.
differently and crudely depicted. All hair styles are wrongly or badly exe-
cuted; some figures seem barefoot, others not. In the second row from the
top in Figure 2 is depicted an Elamite driving an Assyrian chariot (note
the wheel studs), outside of which is a warrior (nationality?) standing on
the axle (?), clubbing an Elamite passenger: this anomaly is set amidst a
nonbattlevictoriousscene (note the pipest serenading the slaughter);
the scene is described without comment (p. 28). Note the full body-covering
shields shown in front view.
In the lower scenes, bowmen fold their hands while also carrying bows;
one bowman is beardless, the others bearded. The groveling Elamites have
unique headbands. In the lowest, fourth, scene (Fig. 1), there is an empty odd
chariot driven by an Elamite (p. 28). To the right, the Elamite kings garment
is wrongly depicted, the side fringe should slope left to right, depictions of
his hat are misunderstood and not uniform and his leg floats in the air; he is
also shown again isolated by the kings chariot.
The kings chariot is depicted twice (Figs. 1 and 4), large and overlapping
two zones, as on the palace reliefs. The royal crown is inconsistently drawn.
The parasol forms are misunderstood. Likewise the fan bearers and fans,
which in one case strike the back of the head of the parasol bearer, in the
other cross behind his head and touch the kings crown; the number of
fan units varies in the two scenes, and the heads of the fan bearers are
inconsistently drawn. The parasol bearers hold neither a parasol handle nor
a fly whisk. The figure walking alongside the chariot wheel in one case grasps
spokes with both hands, politely ignoring the floating pair of bodiless hands
grasping a spoke; on the other chariot, this figure holds a spoke in one hand
while the other hangs at his side; here also in one case one foot is booted,
the other bare. In excavated Assyrian art, this figure holds one spoke, never
two, sometimes also the hub, and in the former position, he carries a baton;
also, his posture is rendered upright, not bent. Behind the lowest chariot
are two figures holding very long, thin, plain poles. Bleibtreu notes (p. 25)
that, bisher (wonderful!in a word disclosing how data are acquired in
bazaar archaeological scholarship), they have no parallels in Assyrian art.
But inasmuch as they are said-to-be ancient artifacts, they are labeled rulers
staffs. Not recognized is that the arrangement of the scenes above bands
of (poorly executed) palmettes is also (bisher?) not Assyrian; nor are the
palmettes Assyrian-they were copied from Phoenician patterns.9
9 See H. Frankforts The Art and Architecture of the Ancient Near East (Harmondsworth
and Baltimore: Penguin, 1954, 1963), fig. 94, and S. Moscati, The World of the Phoenicians
(London: 1968), figs. 10, 19, 20, and pls. 1, 7, 92 (the Moscati reference is from Glenn Markoe).
Fig. 1. Gilt silver beaker, view 1. Height: 24.5cm. Miho Museum, Japan.
Fig. 2. View 2.
Fig. 3. View 3.
Fig. 4. View 4.
Something is very wrong, but not in ancient Assyria. The scenes were
chased,10 consistent with ancient practice; but not mentioned is a corro-
sion analysis of the chasing or inscriptions, or an investigation of the gilding
process. Absent a sense of archaeology and scholarly insight, the question
Why is the beaker ancient? was never asked. To the (anticipated) out-
raged bazaarist defense that one cannot prove by stylistic and iconograph-
ical analyses that the decoration is not ancient, the spontaneous rebound
shoutsand more accurately: one cannot prove that it is ancient. Thus,
whether one accepts modern fraud or ponders an unsure posture, the said-
to-be Assyrian beaker ceases to function in archaeological, cultural, and
art-historical discourse. (Of course, museum and other bazaarist personnel
have their own protocols and will continue to exhibit, publish, and educate
the public on the beakers beauty and uniqueness.)
Consider these scenarios. The beaker is probably ancient (is the base
added?), seemingly one of many plundered from western Iran in the early
1990s and smuggled to western bazaars. It bears a Neo-Elamite inscription,
but, being plain, its sales value is restricted. A bazaar-wise entrepreneur
(known in the bazaar and museum world as an honest and reliable dealer)
employed ayesprovincial artisan (we may call him the Miho Mas-
ter) to embellish the begging-to-be-decorated plain vessel with an Assyrian
inscription and scenes loosely copied from (photographs of?) Assyrian wall
reliefs. As expected (from past experiences), the scheme succeeded: the
Beaker of Ashurbanipal, created to be sold, was sold (of course, in good
faith), exhibited and praised by several museum directors and curators,
baptized (born again) authentic, and published by a believer. Bazaar archae-
ology, sublime and elegant.11 Or both inscriptions as well as the decoration
10 Richard Stone examined the published photo and determined that the incisions were
chased.
11 The Vienna Museum director Wilfried Seipel also proudly exhibited and published as
ancient other Miho forgeries, equally discovered in bazaar stalls by alert directors and schol-
ars. The catalogue entries written by Trudy Kawami (with technische Angabe von Pieter
Meyers) assert as a givenbut never demonstratetheir ancient age: no. 6, Bactrian, and
nos. 12, 13, 14, 15 (utterly perverse). These repeat what Kawami earlier asserted in Ancient Art
from the Shumei Family Collection (no editor!) (New York: Metropolitan Museum of Art, 1996),
nos. 3, 11, 12, 13; no. 10 here is also a forgery, and no. 7 is a modern pastiche. As a welcome anti-
dote to all this, see U. Lw, Figurliche verzierte Metallgefsse aus Nord- und Nordwestiran
(Mnster: Ugarit-Verlag, 1998), pp. 325 ff., 450 ff., 593 (reviewed by me in Archiv fr Orient-
forschung 57, nos. 112 [2000]: 187195), and the Marlik section of my The Lie Became Great
(see n. 2, above).
Forgery factories recruit fairly good (provincial) artisans by guaranteeing them lifetime
jobs.
are recent additions (whether to an ancient or recently made vessel), at-

tempting to associate the beaker with vessels bearing the Iranian name.
Only disinterested laboratory examination by several (nonmuseum-em-
ployed) scientists may resolve the technical issues raised here.
chapter thirty-four
VON BISSINGS MEMPHIS STELA:

A PRODUCT OF CULTURAL TRANSFER?*1
The tracking of cultural transfers begins with evidence of objects that stand
out from the norm. Exotic materials, new technologies and deviations in
form, style or decoration are the tell-tale signs of imports or foreign-inspired
products of local manufacture. Their presence alerts us to the existence of
outside stimuli which raises questions about the nature of contact and its
impact on the recipient culture. Knowing the provenience of these objects,
their archaeological context, is fundamental to the investigation which they
inspire into cultural exchange and interrelations. Without this information,
research is severely hampered but unexcavated objects can still yield valu-
able, if restricted, insights into the nature of outside stimuli, if evaluated in
an appropriate fashion.
In his 1987 article on tracking cultural transfers, Moorey explores the
question of whether writing and other aspects of early civilization reached
Egypt from Mesopotamia along with the import of Uruk-type seals, pottery
and iconography that were found in predynastic contexts. Moorey focused
his research on excavated material, but he was also aware that insights into
the character of cultural exchanges and influences on early state forma-
tion in Egypt stem in part from the evidence of unprovenienced products.
Numerous ivory and embossed sheet-gold knife handles and large carved
slate palettes decorated with figural and floral motifs lack archaeological
context. Deemed authentic and assigned to predynastic Egypt on stylistic
grounds, they bear witness to the syncretistic tastes of an Egyptian elite.
Their selection of certain Near Eastern motifs to decorate native ceremonial
and votive objects supports the general conclusion, based on a wider range
* This chapter originally appeared as Von Bissings Memphis Stela: A Product of Cultural
Transfer? in Culture through Objects: Ancient Near Eastern Studies in Honour of P.R.S. Moorey,
eds. T. Potts, M. Roaf, and D. Stein (Oxford: Griffith Institute, 2003), 109121.
1 It is a personal pleasure and honour for me to dedicate an article to Roger Moorey about
whom many good things can be said, all true. He is a learned and intelligent scholar, he is
utterly honest and generous without hesitation to all, and he is a paradigm of what a scholar
should be in and outside of scholarship.
902 chapter thirty-four
of evidence, that at the end of the fourth millennium bc, elements of foreign
imagery were adopted by an already complex society and adapted to suit its
own ends. Lack of context may restrict a full analysis of these unexcavated
artefacts but provided they are genuine, they contribute to the reconstruc-
tion of intercultural contact.
Different by definition, objects of foreign or multicultural inspiration
sometimes stretch the boundaries of our existing frame of reference. Exca-
vated material is beyond suspicion. But we are obliged to question the
authenticity of unconventional objects that lack provenience and find
viable criteria that will serve to increase our knowledge of the genuine and
eliminate from the record the false information derived from the fake. Fail-
ure to query and investigate the background of such artefacts can lead to a
skewed understanding of the culture in which they allegedly originated. The
present paper is concerned with one of the many examples where scholars
failed to question their sources. To some extent our understanding of Egypts
interaction with areas to the East during the fifth century bc has been embel-
lished by the oversight.
In 1930 Freiherr F.W. von Bissing, an Egyptologist as well as a collector
and vendor of antiquities (Bissing 1930; Anthes 1934: 90), published a stone
relief (called a Totenstele) of yellow limestone (Fig. 1), which he had heard
about from a dealer named Cassira based in Egypt. According to Cassiras
testimony, presented by von Bissing with conviction, the stela was discov-
ered at Mithrine, the site of ancient Memphis, in 1909. Von Bissing neglects
to mention in his 1930 publication that he owned the stela or when he pur-
chased it from Cassira (cf. Parlasca 1972, below).2 The important fact of the
stelas purchase and its implications were therefore consistently overlooked
by scholars who cited the relief in subsequent years, accepting it as having
derived from Memphis. Indeed, some took for granted that it was excavated
there, resulting in the name the Memphis Stela by which the object has come
to be known in literature (viz. Anthes 1934; C. Picard 1935; Culican 1965: 154;
Kyrieleis 1969: 39, 148; Parlasca 1972: 76; Gallo 1993: 273; Martin and Nicholls
1978: 6667; Ray 1988: 273; Briant 1992: 9091; 1996: 974; Mathieson et al.
1995).
2 When did he purchase the relief? In the photo caption for the piece (Abb. 1), its
provenance is given as Museum Carnelielaan, den Haag (which fooled me in Muscarella

1980: 26), but nothing in the text explains this caption. Von Bissing (1931: 15) acknowledges
ownership of this relief along with another one he claims also came from Memphis. Cassira
is mentioned only obliquely (ibid. 17); no date is supplied (see n. 5).
von bissings memphis stela: cultural transfer? 903
Fig. 1. Memphis Stela (Egyptian Museum, Berlin).
The Memphis Stela is rectangular in shape, measuring 23 cm in height,

45cm in length and 8cm in width. Von Bissing described its scenes in some
detail (1930: 226227): the funeral bed, the Taburett at its side, the pair of
figures at each end, the horizontal projections on either side that function
both as a roof over the lower figures and as a base for the smaller figures
at the top corners. The relief was originally painted with traces of reddish-
brown colour still preserved on the face and arms of the deceased, and at
the time of purchase, there were also traces of black on his beard.3
The corpse is described as lying on three mattresses and the rectangular
mass above the bed is thought to be a continuation of the cover that passes
under the body and hangs behind the table or stool. The deceased wears
a sleeved garment and perhaps shoes. His beard reaches to his breast. His
hair bulges at the brow and twisted locks reach his shoulders leaving the ear
free. Von Bissing likened the hat to a beret and described the attachments
as bands or side flaps that hang down over the second mattress (cf. Bissing
1930: 232).4
At the head of the funeral bed are two mourners, both bare-breasted
females according to von Bissing because of the bulge on the chest of the rear
figure and the mass at their waists which he interprets as rolled-up under-
clothing. The left figure caresses the dead mans head, a typical Egyptian
3 Von Bissings reference to the time of purchase could be understood as an oblique
reference to his acquisition of the Memphis Stela.

4 In fact, this seems not to be the case, for the flap appears to hang from the back of the
hat.
gesture according to von Bissing, while the right figure tears at her hair. At
the foot of the bed stand two men, apparently without shoes, each wearing
a knee-length garment over trousers as well as a hat with an appendage very
much like that on the hat of the deceased except that here it clearly hangs
from the back. The two figures visible in the upper corners are described as
sirens without further explanation (ibid. 229), possibly because the one at
the right seems to have a feathered tail and a clawed front foot. The smaller
one to the left overlaps the bed cover to leave room for an open-mouthed
horse that is led by a servant with Egyptian hair. He wears a jacket but no
trousers, hat or shoes.
Von Bissing noted that his relief was strange and its style coarse (ibid.
229), but as he never doubted its authenticity, these observations did not
deter him from attempting to find its place within the artistic traditions
of the ancient world. As a relevant comparison, in particular for the pres-
ence of mourners in the death scene, he cites a fifth-century bc Egyptian
stela with an Aramaic inscription in the Vatican (ibid. 229230, Fig. 3). The
only parallel for the funeral bed with its upturned head section as well as
for the stool comes from Neo-Assyrian art; namely, Assurbanipals garden
relief (ibid. 231). Also considered Assyrian in inspiration are the hair and
beard of the deceased. Von Bissing refers to a statue of Assurnasirpal II and
reliefs of Sargon II from the ninth and eighth centuries bc although he also
notes deviations. The typically Assyrian horizontal division of the beard, for
example, does not occur on the Memphis relief; a deviation which von Biss-
ing explains as a misunderstanding on the part of the ancient craftsman and
construes as evidence for its Egyptian manufacture. Another comparison for
the hair and the dead mans hat with what he saw as side flaps leads him to
a vessel from the Crimea (ibid. 231, Fig. 4) which appears to show Persians
wearing hats with side flapsunlike those worn by the Persian tribes on the
Persepolis reliefs that have only one appendage hanging from the back. The
hair of these Persians also compares favourably with that of the deceased
and his attendants on the Memphis relief.
The funeral bed, the stool as well as the dead mans hair and beard all
suggested an Assyrian background. However, the hats with flaps pointed to
a Persian connection, as did the trousers worn by the male mourners on the
Memphis Stela and the presence of a horse which played an important role
in Persia (ibid. 234). The Memphis Stela also included Greek and Egyptian
elements. The sirens were deemed to be Greek, although von Bissing knew of
no parallels (ibid. 235), and the wailing women recalled Egyptian practices,
although it appears that they were not unknown among Persians (ibid. 231
232).
Faced with a stela that was said to come from Memphis and had obvious
Assyro-Persian comparanda, von Bissing concluded that the dead man was
an important Persian resident in Egypt who died during the reign of Darius
the Great, at the end of the sixth century bc, when Achaemenid art still
closely adhered to Assyrian prototypes. To him, such works incorporating
Persian, Greek and Egyptian elements were not uncommon in Memphis
and he presented as evidence other examples, a number of hybrid seal
impressions and terracotta figurines of foreigners, all of which were also
unprovenienced and privately owned, indeed, some by von Bissing himself
(ibid. 234235).5
R. Anthes published the stela in 1934, recording its purchase that year by
the Berlin Museum (no. 23721) from von Bissing. He did not doubt that it
derived from Memphis (1934: 99). This findspot, the material, the coarse
technical execution and the overall style were all indications, he claimed,
of its Egyptian origin, although the rest appeared to him to be strange. The
shape has no Egyptian parallels, although he calls it a stela, and he notes
that the representation of the dead mans right arm rising from behind the
body is also un-Egyptian. Anthes refers to the rectangular area over the dead
mans head as a Bahrtuch and offers only Greek parallels (ibid. 93, n. 2). Of
the figures, only the mourning females could be considered Egyptian and
perhaps also the servant beside the horse, unless he wears the same clothes
as the male figures below. Anthes does not cite the Vatican stela (ibid. 100)
but agrees with von Bissing that the bed and table and the hair and beard are
Assyrian in style, and that the clothing worn by the dead man and his two
male attendants recall the attire of the Persians and Scythians represented
at Persepolis. Their hats, however, he would not compare with those on the
Crimean vessel although they could fit into an Iranian or Scythian sphere;
to support this view he cites the horse. He too cannot find parallels for the
sirens. Thus, non-Egyptian (that is, Assyrian and vaguely Iranian) elements
5 Von Bissing also published four other photographs at the end of his text, only two of
which he mentions in the text (1930: 236). One is his Abb. 6, a clay head, which he says is
Achaemenid from Memphis. He compares it with a clay head (Abb. 8) which he says is a
Semite (no provenience is given). Both would appear to be forgeries. In the photo credits
on p. 238 it transpires that Abb. 5, a clay horseman, Abb. 7, a clay head, and Abb. 8 are in
his private collection. The last two are listed in the photo captions as from Memphis and,
again, both are surely forgeries. The photo credits fail to mention that Abb. 1, the relief under
discussion here, also belonged to von Bissing. I thank Professor Wildung of the Egyptian
Museum in Berlin for granting me permission to publish this relief, knowing that I question
its authenticity.
prevail. Because of the lack of Persian elements, Anthes favours an earlier

date than von Bissing, placing it before the Achaemenid period, in the Saite
Dynasty, 663525bc. For him too, the dead man was a foreigner who died
and was buried in Egypt (ibid. 101).
C. Picard (1935) accepted that the relief, here now correctly called a pan-
neau rather than a stela,6 was made of Memphite stone. Like Anthes, he
too recognized that the shape of the relief is unknown in Egypt, but he
ignored Anthes disclaimer about the visible rear arm and refers only to the
portrayal of one arm which he considers to be conventional (ibid. 93). He
compares the scene to that of a Greek prothesis but questions the Greek
origin of the sirens, citing for comparison instead a famous relief from Xan-
thos in the British Museum (ibid. 93). The female mourners may be Egyp-
tian, as von Bissing and Anthes claimed, but Picard points out that they
have close analogues on the Phoenician sarcophagus of Ahiram and that
the custom of wailing was not confined to Egypt. In fact, he thinks that
the model of death is Iranian. In common with his predecessors, Picard
recognizes isolated Assyrian elements such as hair style. The horse, he
says, belonged to a satrap and is evidence of the syncretistic nature of the
relief, which he considers to be a rare et prcieux witness to the pres-
ence of Persians in Egypt. Picard disagrees with Anthes early dating, believ-
ing the foreign resident to have lived at the time of Cambyses (ibid. 93
94).7
W. Culican thought the representation from Memphis to be curious,
depicting the funeral of a Persian or Median noble (1965: 154155). To
him, the bed and table resemble Assyrian- or Achaemenid-style furniture,
likewise for the style of the dead mans hair and beard. However, the dead
mans sleeved garment and hat are Median, the male mourners at the foot
of the bed are Median or Scyths and the presence of a horse at a burial is a
Scythian custom.
H. Kyrieleis, referring only casually to the Memphis grave relief with-
out regard to previous publications (1969: 19), states that it was a Persian-
Egyptian relief from Memphis of the fifth century bc. In keeping with his
general discussion of furniture, he did not discuss the representations on
the relief except to point out (like von Bissing before him) that the best
6 Scholars subsequently referred to it either as a stela or a relief.

7 Picard claimed that both von Bissing and Anthes dated the relief to the period of
Psammetichus. In fact, only Anthes dated the relief that early. Thus, Picard actually agreed
with von Bissing.
parallel for the funeral bed was that on Assurbanipals garden relief. To
Kyrieleis, the rectangular area over the funeral bed is a cloth characteristic of
Persian custom although in support of this claim, he cited not a Persian bed
but the cloth covering Persian throne seats (ibid. 148). K. Parlasca accorded
the relief only one sentence and a footnote to observe that it derived from
Memphis and depicted the laying-out of a Persian from the time around
500bc (1972: 76 with n. 22). He repeated that the relief was found in 1909
and that von Bissing purchased it from Cassira in Cairo, not previously
mentioned by von Bissing. When he next referred to the relief, he described
it as among the most famous archaeological testimonies of the first Persian
period (1979: 318).
In more recent times von Bissings stela continued to be accepted as an
important Egyptian monument from Memphis that merited serious dis-
cussion. According to G.T. Martin and R.V. Nicholls (1978: 6667), it is a
stela that represents a Persian dignitary in Median dress from Memphis.
They note comparisons with the Persepolis reliefs but also bring in
Graeco-Persian grave reliefs from Daskylion and elsewhere in Asia Minor
and conclude that it shows an essentially Greek concept rendered in a
Graeco-Persian style. The relief is considered late sixth century bc in date
because in their view the deceaseds beard and Median dress and the
curving profile of the horses head reflect the conventions developed by
Darius at Persepolis. The authors relate the reliefs style and technique to
the Vatican stela which, although unexcavated, they also believed originated
from Memphis.
In 1980 (36) I too cited the piece, not as a forgery, for I barely looked at it,
but as a forgery of provenience because there was no evidence that it derived
from an Egyptian site.
In the Cambridge Ancient History, J.D. Ray refers to the stela to demon-
strate that Egypt was a richly cosmopolitan state (1988: 273). Again, the
scene is described as showing a foreigner in Median costume but Ray
ventures further to propose an historical interpretation, suggesting that the
representation could be taken as a likeness of Atiyawahi or his brother
(Egyptian children of a recorded Persian-Egyptian marriage) and that, more-
over, the artist may have been East Greek.
P. Briant accepts that the subject of the Memphis relief is a Persian noble,
but following Picard, he compares the scene with that of a Greek prothesis
(1992: 9091). Later, he refers to the same relief as an intressant document
funraire that was found at Memphis and portrays a man clothed as a
Persian or Mede (1996: 974). Gallo, in his article on stelae, merely mentions
the relief as being from Memphis and depicting the prothesis of a Persian
(1993: 273). Finally, we find reference to the Memphis Stela in an article by

Mathieson et al. (1995: 38). Taking for granted that it comes from Memphis
and that it shows the mortuary rites of a Persian, they cite the stela as a
parallel to an excavated example (see below) and with that they confer upon
the Memphis Stela the ultimate stamp of approval.
By following von Bissings lead in accepting a dealers parti pris state-
ment about the source of an item on sale, scholars have treated the Memphis
Stela as an excavated object that needs only to be assigned an ancient date
to become a useful index of intercultural connections. The dealers claim
became fact and rendered further investigation into anomalies unneces-
sary; an example of blind faith that remains quite common in archaeological
research of the 20th century.8 In this particular case, failure to check the
reliability of the source and to question aberrations in shape, content and
design have affected our perception of Perso-Egyptian interaction and cul-
tural transfer.
Evidence for cross-cultural influence between Egypt and Persia during
the fifth century bc is well attested in the written and visual records of the
Achaemenid empire. Recent interpretations of Achaemenid art as exempli-
fied, for example, in the writings of Margaret Root (1979), propose that offi-
cial Achaemenid art was used as a vehicle for propaganda and that images
from the court were widely dispersed over the entire empire. A version
of the Bisitun relief has been reconstructed from carved stone, and pos-
sibly also from glazed brick, fragments found in Babylon and Susa (Seidl
1976, 1999; Canby 1979; Muscarella 1992: 221, n. 14), and it has been sug-
gested that similar pictorial monuments were set up by Darius through-
out the Persian empire just as copies of the Bisitun inscription were sent
out and indeed have been found in Babylon (Seidl 1999) and Elephantine
in Egypt. Achaemenid royal imagery was known in the empire and doc-
umented abroad. The scene of the enthroned king, for example, which
formed the centre piece of the Apadana faades at Persepolis and the main
motif of the doorways of the Hall of 100 Columns, is found on the inner side
of a shield carried by a Persian on the Alexander sarcophagus from Sidon
and on seal impressions from Daskyleion (Root 1979: 12, 122).
There are also monuments in which Egyptian and Persian conven-
tions are combined. The Canal Stelae of Darius, for example, have one side
8 For the predisposition of scholars to equate dealers claims about proveniences with
archaeological reports, see Muscarella 2000a: 1314, passim. See also idem (2000b: 2937).
inscribed in cuneiform and the other in Egyptian hieroglyphics (Root 1979:

6162). Another example is the Statue of Darius found in 1972 at Susa but
undoubtedly carved in Egypt from Egyptian stone (Root 1979: 6869; Mus-
carella 1992: 219220, Fig. 50, with bibliography). Its base is essentially Egyp-
tian but traces of multicultural influence are evident in the un-Egyptian
representation of the subject peoples, whose hands are raised in prayer
or supplication rather than tied behind their backs, and in the misinter-
pretation of certain motifs such as the bell-shaped headdress worn by the
Babylonian which the Egyptian sculptor rendered as the White Crown of
Egypt. The statue of the king follows almost exactly the models of the reliefs
at Persepolis albeit with a few traditional Egyptian elements such as the
back pillar and the short staff held in the closed fist. Whether these Egyp-
tian features would have been typical of Achaemenid royal statues made
outside Egypt, like other Egyptian features such as the winged disc and
cavetto cornice that appear at Persepolis, cannot be confirmed, for apart
from this statue no monumental sculpture of the Persian king has been
recovered anywhere. There are five different coloured stone fragments also
from Susa (Root 1979: 111112; Muscarella 1992: 219220) that depict at least
two to four human figures, one or more of which might have come from a
royal statue, but their place of manufacture, whether in Susa or in Egypt,
remains unknown.
The extent to which Achaemenid court art influenced the non-royal or
unofficial art produced in Egypt is difficult to establish. There were Egyp-
tians present in Persia (as attested by the Persepolis Fortification Texts) and
Persian and other foreigners in the service of the Persian administration
were resident in Egypt. Sometimes there is evidence for marriages between
Persians and Egyptians but how much the various different ethnic commu-
nities influenced each other is in most cases uncertain.
Speculation about the adoption of Egyptian mortuary practices in Per-
sia was recently revived by a said to have been discovered Persian Princess
whose mummified body surfaced in Pakistan. The inscribed wood sarcoph-
agus containing the mummy was claimed to belong to a daughter of Xerxes.
The entire unit (mummy, sarcophagus and inscription) is in fact a badly con-
structed forgery. Nor is there any substance to the claim that Persians prac-
ticed embalming in their homeland as Herodotus (I, 140) was incorrectly
alleged to have reported; in fact, he states only that they covered their dead
with wax before burial. That someone should devise such a ruse, attempt
to market it and even succeed in fooling some individuals, suggests that the
forgers believed there was a case for documenting that Egyptian practices
were adopted by the Persians. Here, of course, the Memphis Stela comes to
mind.9 but as a purchased item, it requires careful scrutiny before it can be

assimilated into the debate on Persian residents abroad.
For comparison and as an objective control on the stylistic criteria used
to evaluate the Memphis Stela, we now have the advantage of reference to
a similar syncretistic scene, this one on an excavated funerary stela from
Saqqara (Fig. 2), the ancient cemetery at Memphis (Mathieson et al. 1995:
26ff., Fig. 3). Like its genuine counterpart from Saqqara, von Bissings stela is
made of limestone;10 however, its shape and state of preservation differ. The
Saqqara stela is conventional in form with a rectangular body surmounted
by a rounded top (like the Vatican stela mentioned above) and apart from
the natural effects of erosion and minor damages to its edges, is in excellent
condition (Mathieson et al. 1995: 26). The rectangular shape of von Bissings
relief struck many scholars as odd and, as Picard first pointed out, it has more
in common with a relief. The evidence against accepting von Bissings relief
as ancient increases when we compare its subject matter and iconography
with those of the Saqqara stela. Both monuments depict the burial rites
of a prominent foreigner. In the case of the Saqqara stela, the inscription
records that the deceased was of mixed Persian and Egyptian parentage
and the iconography of a mummy and an enthroned figure in Persian dress
(whoever he may be) makes clear that the family was both well off and well
connected. Von Bissings purchased relief is not inscribed and the deceaseds
identification as a Persian of high standing has been assumed from his
beard, attire, furniture, attendant mourners and the horse.
As hybrid products of Egypto-Persian background, one expects both
monuments to include elements of local (Egyptian) and foreign (Persian)
iconography. Nor should the presence of unparalleled traits surprise us a
conventions of one culture were often misunderstood by craftsmen from
another and, in any case, our perception of the norm is liable to change. On
the Saqqara stela only a minority of Persian features appear to be unique: the
9 I became interested in the Memphis relief after my involvement with the so-called Per-
sian Princess, which I determined was a forgery. Because some individuals, including several
Zoroastrians in California and several newspapers, asserted that the mummy documented
Persian practice of mummification after the Egyptian fashion, I decided to investigate. It was
then that I recalled the von Bissing relief in Berlin and began to look into its background
(more fully than I did for my 1980 paper!). For the true history and uncovering of the forgery
see Romey and Rose (2001). The creators of a TV documentary about the mummy give an
inaccurate account.
10 The Saqqara stela is made of Tura limestone. That of the Memphis Stela is described a
yellow but has not been analyzed.

Fig. 2. Saqqara stela (Mathieson et al. 1995: Fig. 3).
circlet around the head of the enthroned Persian, the drape and mattress
on the chair on which he sits, the torque or ring presented to him,
the second offering table and the finger position which seems closer to the
Assyrian manner (Mathieson et al. 1995: 3031). All other elements of
the funerary scene can be traced in Late Period Egyptian or Achaemenid art
while some, like the winged disk and the lion bier, combine characteristics
of both (ibid. 28, 31).
On von Bissings stela viable and recognizable elements and compar-
isons are few and imprecisely rendered. The two sirens and the rectan-
gular bed cover have no parallels. The stool/table, although considered
Assyrian, is also without analogues and like other sections of the relief
seems to have been consciously mutilated to suggest age or conceal poor
workmanship.11 The funerary bed is said to resemble the furniture in Assur-

banipals garden relief, but note the awkward rendering of the headrest, on
which the deceaseds head rests improperly, and the two differently carved
feet. The dead mans hair and beard are fashioned after Assyrian style but
with marked differences. His raised rear arm conforms to neither Egyp-
tian nor Persian convention. His hat and those of his male attendants are
alternately described as Persian, Scythian or Median. Unlike Scythians and
Medes, these men wear tight trousers and their shoes are not adorned with
straps. Egyptian and Phoenician parallels have been cited for the two female
mourners but there are distinct formal and stylistic differences in posture,
clothing and arm positions, notably in the case of the second mourner
whose right hand appears unattached above her head. Its large proportion
of incomparable traits, lack of provenience, unusual shape and inexplicable
damage (see below) all argue against von Bissings stela. In addition there is
the anomaly of craftsmanship. Described as coarse by its owner, the quality
and style of von Bissings stela do not compare with the careful modeling of
the Saqqara stela. While subjective and normally oflittle consequence, this
observation does have a bearing on the case at hand when one considers
that the deceased in both cases appear to come from wealthy families. The
dead man on von Bissings stela had expensive furniture, attendants and a
horse. It is unlikely that his family provided him with an elaborate funeral
but could not afford an experienced craftsman or a stone that was better
suited to a multi-level composition. The Memphis Stela was probably made
in Egypt by a craftsman who lived in the early 20th century.
In the current climate of archaeological research, which advocates the
use of only verifiable sources of information, objects that come from con-
trolled excavations take precedence over those that do not (Muscarella
2000a). This privileging of excavated material ensures a sound database
for research and inevitably casts suspicion on all objects of unknown ori-
gin, both genuine and fake. Scholars who want to learn information from
unprovenienced objects are obliged to make a case for accepting them
as manifestly ancient rather than assuming this on misplaced faith alone.
The verdict on the case of von Bissings relief paradigms the problem. For
11 Deliberate damage to reliefs is usually the result of a politically motivated act of
iconoclasm directed at key figures of authority. The funerary monument of a nameless foreign
resident in Egypt is unlikely to have aroused such feelings of aggression and, in any case, the
damage to the Memphis relief is randomly distributed, affecting not only the face and left
hand of the deceased but two of his attendant mourners and a harmless stool/table as well.
discussions on Persian-Egyptian relations and shared customs in the fifth

century bc, the remarkable funerary stela excavated at Saqqara, not the von
Bissing relief, is a source.12
References Cited
Anthes, R.
1934 Fnf Neuerwerbungen in der gyptischen Abteilung. Berliner Museen Be-
richte aus den Preussischen Sammlungen 55: 90101.
Bissing, F.W. von

1930 Totenstele eines persischen Groen aus Memphis. Zeitschrift der Deutschen
Morgenlndischen Gesellsthaft NF9: 226238.
1931 Osiris im Boot. Zeitschrift fr gyptische Sprache und Altertumskunde 67:
115119.
Briant, P.
1992 Darius. Les Perses et lempire. Gallimard.
1996 Histoire de lempire perse. De Cyrus Alexandre. Paris: Fayard.
Canby, J.V.
1979 A note on some Susa bricks. Archologische Mitteilungen aus Iran 12: 315
320.
Culican, W.
1965 The Medes and the Persians. New York.
Gallo, P.
1993 Une stle Hellnomemphite de lex-collection Nahhman. Bulletin de lin-
stitut franais darchologie orientale 93: 265274.
Kyrieleis, H.
1969 Throne und Kline. Berlin.
Martin, G.T. and Nicholls, R.V.

1978 untitled contribution in Oliver Mason, Carian Inscriptions from North Saq-
qra and Buhen. London: Egypt Exploration Society, 5787.
12 Although not funerary scenes, the Petosiris reliefs from Egypt (of post-Achaemenid
date) are also relevant for such discussions. Here we have classic Egyptian-made tomb reliefs
with scenes of ethnic Egyptian craftsmen accurately making Achaemenid-style vessels. See
Muscarella (1980: 2829, Pls. VIIIIX).
Mathieson, I. et al.
1995 A stela of the Persian period from Saqqara. Journal of Egyptian Archaeology
81: 2341.
Moorey, P.R.S.
1987 On tracking cultural transfers in prehistory: the case of Egypt and lower
Mesopotamia in the fourth millennium bc. In Centre and Periphery in the
Ancient World (M. Rowlands, M. Larsen and K. Kristiansen, eds.). Cam-
bridge: Cambridge University Press, 3646.
Muscarella, O.W.
1980 Excavated and unexcavated Achaemenian art. In Ancient Persia: the Art of
an Empire (D. Schmandt-Besserat, ed.). Malibu, California: Undena Press,
2342.
1992 Fragment of a royal head. In The Royal City of Susa (P.O. Harper et al., eds.).
New York: Metropolitan Museum of Art, 219221.
2000a The Lie Became Great. Groningen: STYX Publications.
2000b Excavated in the Bazaar: Ashurnasirpals beaker. Source 20/1: 2937.
Parlasca, K.
1972 Eine Gruppe rmische Sepulkralreliefs aus gypten. Forshungen und Be-
richte 14: 7278.
1979 Persische Elemente in der Frhptolemischen Kunst, Akten des VII. inter-
nationalen Kongresses fr iranische Kunst und Archologie, Munich, Sep-
tember 710, 1976. Archologische Mitteilungen aus Iran, Ergnzungsband
6: 317323.
Picard, C.
1935 La satrape au tombeau, relief de Memphis. Revue Archologique 6: 9294.
Ray, J.D.
1988 Egypt 525404bc. In Cambridge Ancient History 4/2, 254286.
Romey K.M. and Rose, M.

2001 Saga of the Persian princess. Archaeology January/February: 2425.
Root, M.
1979 The King and Kingship in Achaemenid Art. Acta Iranica 19, Series 3, Leiden.
Seidl, U.
1976 Ein Relief Darius I. in Babylon. Arcologische Mitteilungen aus Iran 9: 125
130.
1999 Ein Monument Darius I. aus Babylon. Zeitschrift fr Assyriologie und Arch-
ologie 89/1: 101114.
chapter thirty-five
GUDEA OR NOT GUDEA IN NEW YORK AND DETROIT:

ANCIENT OR MODERN?*
Cultural problems to be resolved by the methodology that art historians

call connoisseurship are also addressed by archaeologists and other schol-
ars concerned with ancient Near Eastern art and artifacts. Although the
term connoisseurship is rarely used in archaeologystylistic analysis or
art-historical analysis are more common archaeological termsthe basic
methodology is the same. But connoisseurship is particularly relevant to
archaeology in cases of objects that surface not from a controlled and re-
corded excavation, but from a dealers shopthe antiquities market. In
these cases, the object has no provenience, meaning that we possess no
knowledge of its original source. As a result, we know nothing of its histor-
ical and cultural context, and the first question for scholars is whether the
archaeologically undocumented object is in fact ancient or a forgery. The
answers are not always straightforward. The ancientversus-forgery connois-
seurship issue considered here is but one example of the problems inher-
ent in attributions that occur when an archaeologist confronts unexcavated
objects that are unique, odd, or aberrant.
Two statues present an interesting connoisseurship attribution chal-
lenge. Both are in museum collections, both were purchased, and their
proveniencesi.e., their archaeological lociare unknown. The first
statue, carved in diorite, is in the Metropolitan Museum of Art (Fig. 1a and
1b) and represents a seated Gudea, the governor (ensi) of the Sumerian state
of Lagash in Mesopotamia during the late third millennium bc. The sec-
ond statue, also claimed to represent Gudea (Fig. 2a and 2b) and owned by
the Detroit Institute of Arts, shows him standing and is carved in the green
stone paragonite. The difficulty is that the Metropolitan Museums Gudea
conforms to a group of documented statues, but its inscription is problem-
atic, whereas the Detroit statue conforms to no other known Gudea statue,
but some cuneiform scholars have deemed its inscription authentic. This is
the connoisseurship problem that I address.
* This article originally appeared as Gudea or Not Gudea in New York and Detroit:
Ancient or Modern? Source 24, no. 2 (2004): 68.

916 chapter thirty-five
Fig. 1a. Gudea. Diorite. The Metropolitan Museum of Art, Harris Brisbane
Dick Fund, 1959 (59.2). Image The Metropolitan Museum of Art.
gudea or not gudea in new york and detroit 917
Fig. 1b. Side view of Figure 1a.

Fig. 2a. Gudea. Paragonite. Detroit Fig. 2b. Three-quarter view

Institute of Arts, Founders of the chest, shoulders,
Society, Robert H. Tananahil and head of Figure 2a.
Foundation Fund, F82.64.
The Background
During two extended campaigns in southern Iraq (1877/18781909 and 1929

1933), French archaeologists excavated mounds at Tello, known from in-
scriptions to be the city of Girsu in Lagash.1 The 18801881 campaign yielded
a group of seated and standing statues of Gudea. According to the excavation
director, Ernst de Sarzec, they were recovered in a Hellenistic palace dating
to the late first millennium bc. Why and how the statues were recovered in
a context dated almost two millennia after their creation is archaeologically
intriguing.2 Further, beginning in the mid-1920s a number of Gudea statues
began to be offered for sale by European antiquities dealers, some of which
were claimed to have derived from Tello in 1924.3
Eleven or twelve statues (plus fragments of others) were recorded as
excavated from Tello; they were labeled and identified by alphabet letters.4
Seven examples are of standing figures; five of them are in the Louvre,
all headless (for example, Fig. 3, statue A), one is in London, and one is
reconstructed from fragments housed in the Istanbul Museum and the
1 Some of the issues mentioned here have been discussed in less detail in Oscar White
Muscarella, Seated Statue of Gudea, in Art of the First Cities: The Third Millennium bc from the
Mediterranean to the Indus, ed. J. Aruz and R. Wallenfels (New York: Metropolitan Museum
of Art, 2003), pp. 428430, no. 305.
2 Questions had been raised about whether the statues had actually been excavated
in the temple or purchased from local Arabs who had previously plundered the site. See
A. Parrot, Tello: Vingt Campagnes de Fouilles (18771933) (Paris: 1948), pp. 1516 and n. 9,
p. 18, and p. 152; F. Johansen, Statues of Gudea, Ancient and Modern (Copenhagen: 1978),
p. 29 and n. 104; and D.P. Hansen, A Sculpture of Gudea, Govemor of Lagash, Bulletin of
the Detroit Institute of Arts 64, no. 1 (1988):18, n. 1. Parrot expresses his suspicions by writing
(pp. 160164) Trouve par de Sarzec au Palais (7), whereas in his reference to a statue
excavated in 1903 (p. 165, no. 13) there is no question mark. I. Winter, Idols of the King:
Royal Images as Recipients of Ritual Action in Ancient Mesopotamia, Journal of Ritual
Studies 6, no. 1 (1992):18, and H. Vogel, Statuen, die sichtbar machen zur Bedeutung der
Statuen des Gudea, Baghdader Mitteilungen 31 (2000):17, n. 7, accept the Seleucid-palace
locus.
3 Gudea statues were first offered for sale in 1925. For the 1924 plunder, see, among
others, Parrot, pp. 27, 167; Johansen, pp. 56, 1819, 32, who doubts the date because it was a
dealer who reported it (see n. 8, below); E. Mller, On the Gudea Statue in the NY Carlsberg
Glyptotek, Assyriological Miscellanies (1980):5253; also nn. 810, below.
4 L.A. Heuzey, Catalogue des antiquits chaldennes: Sculpture et gravure la pointe (Paris:
1902), pp. 167188; Parrot, pp. 160172; E. Strommenger, Gudea, Reallexikon der Assyriolo-
gie und vorderasiatischen Archologie (Berlin: 1971), pp. 681682; and E.A. Braun-Holzinger,
Mesopotamische Weihgaben der frhdynastischen bis altbabylonischen Zeit (Heidelberg: 1991),
pp. 158159, all provide lists of the excavated and not excavated statues. See also Muscarella,
p. 429 and n. 4.
Fig. 3. Statue A (Gudea) from Tello. Louvre, Paris, AO 8.

Louvre.5 There are also five seated figures in the Louvre (for instance, Fig. 4,
statue I), four of them headless.6 Two isolated heads in the Louvre also derive
from Tello, one comes from Nippur, and three come from Ur.7
All the dealer-derived, unexcavated Gudea statues are in museum collec-
tions; seven are standing,8 and three (possibly four) are seated.9 There are
also about sixteen isolated Gudea heads, more than twice the number of
excavated examples.10
From ancient texts, we know that Gudea built many temples and was
engaged in trade with many distant lands from which he acquired precious
5 The standing statues were identified and labeled as A, C, E, G, K, S, and U. Johansen,
pp. 713, pls. 116; Muscarella Seated Statue, p. 429, n. 5. Statue L is rejected as that of Gudea
by Braun-Holzinger (p. 265), H. Steible (Versuch einer Chronologie der Statuen von Lagash,
Mitteilungen der deutschen Orient-Gesellschaft 126 [1994]:83), and C.E. Suter (Gudeas Temple
Building: The Representation of an Early Mesopotamian Ruler in Text and Image [Groningen:
2000], pp. 330, 335) and is not counted here. U is a fragment in the British Museum that was
excavated in 1850 by W. Loftus at Tell Hamman in southern Mesopotamia.
6 For statues B, D, F, H, and I, see Parrot, pp. 161165, nos. 2, 4, 6, 8, 13; Johansen, pp. 1012,
pls. 1943; and Strommenger, p. 682, 1b 15. Statue I is completely preserved (see statue P,
below); but note that most of the excavated figures are headless.
7 Parrot, pp. 168170, nos. 21, 22 (the latter is bald), 25, and pp. 165, 170171, nos. 12, 27
32; C.L. Woolley, The Early Periods: A Report on the Sites and Objects prior in Date to the Third
Dynasty of Ur Discovered in the Course of the Excavations, Ur Excavations, vol. 4 (Philadelphia:
1955), pp. 51, 189, pl. 42; and Suter, pp. 332333. Most of the excavated figures are headless.
8 Copenhagens Nationalmuseet (O), two in the Louvre (N), British Museum (V), the
Cleveland Museum of Art, the Detroit Institute of Arts (M); and possibly an example in the
Golnisev Collection (Braun-Holzinger, p. 267, no. 124, T [I have not seen it]). See Muscarella,
p. 429 and n. 9; most of these have heads. For scholarly discussions about possible forgeries
among the unexcavated corpus of Gudea statues and heads, see Oscar White Muscarella, The
Lie Became Great: The Forgery of Ancient Near Eastern Cultures (Groningen: 2000), pp. 172173.
9 The Metropolitan Museum of Art (P), Baghdads National Museum (Q), and the British
Museum (acquired in the mid-nineteenth century from Mesopotamia) may not be Gudea;
of these, only the Metropolitan Museum statue preserves its head (see Muscarella, Seated
Statue, p. 429 and n. 10). A damaged statue in Harvards Semitic Museum (8825; new No.
1936.5.1), called R, may not be a statue of Gudea. E. Sollberger, Selected Texts from American
Collections, Journal of Cuneiform Studies 11 (1956):1113, fig. 1, claims the inscription was not
made at Gudeas order, although it may depict him; he also calls attention to the queer
attitude of the torso leaning to its right. Johansen, pp. 25, 40, and Steible, 83, also do not
consider it to be a statue of Gudea. Parrot, p. 171, published it as Gudea, no. 31, pl. XVII, as
does Strommenger, pp. 682683, d, 1.
10 They are housed in New Yorks Metropolitan Museum of Art; Bostons Museum of Fine
Arts; the Cleveland Museum of Art; Harvards Fogg Art Museum; Harvards Semitic Museum;
University Museum of the University of Pennsylvania; Norfolk, Virginia, Chrysler Museum
of Art; Pariss Louvre; Berlins Staatliche Museen; Leidens Rijksmuseum; Jerusalems Israel
Museum; the National Gallery of Victoria, Australia, and Madrids Museo Nacional del Prado.
See Muscarella, Seated Statue, p. 429, n. 11.
Fig. 4. Statue I (Gudea) from Tello. Louvre, Paris, AO 3293.

materials not available in Sumer, such as dark stone.11 Continuing a long

tradition, Gudea commissioned statues of himself that were understood to
embody his living presence. Typically, these statues are carved in diorite
and depict a beardless man whose wide-eyed gaze seems calm and focused.
And they bear an inscription proclaiming his piety and deeds, along with
the complete personal name of his statue. All the figures are dressed in an
ankle-length garment that leaves the right shoulder bare, and they usually
wear a thick-brimmed hat with carved curls that may represent fur (a few
bareheaded examples exist as well). The feet are bare, the arms folded and
subtly muscled, and the hands joined above the waist with the right folded
into the left, an attitude expressing subtle power and control.12
The Connoisseurship Problem
The Metropolitan Museums seated Gudea, labeled P in the corpus inventory

(Fig. 1a and 1b; MMA 59.2) resembles the Louvres. excavated Tello examples
in form and proportion.13 Its closest parallel from this group (its twin
they are within 1 centimeter of each other in height) is statue I (Fig. 4).
Both have a complete head and body, they are the smallest of the seated
statues, and their inscriptions are almost identical. However, the crucial
written (long) personal name of statue P (Let the life of Gudea, who built
the house, be long) is identical to that on the Tello standing statue, labeled
C, in the Louvre.14
In his monograph on Gudea statues, F. Johansen declared statue P to be
a modern forgery: to him it is very much the same in form as statue Iit is
11 See the inscription on statue B in D. O, Edzard, Gudea and His Dynasty, The Royal
Inscriptions of Mesopotamia, Early Periods, vol. 3/1 (Toronto: 1997), pp. 3136. See also Suter
for a good discussion of the political, textual, and material cultural history and records of the
Lagash II and Ur III periods, especially of Gudeas time, and J. Evans, Approaching the Divine
, in Aruz and Wallenfels, pp. 417424, for a neat, concise summary of the period.
12 For a bibliography on the Sumerian and Akkadian background to clothing, hand posi-
tion, named statues, etc., see Muscarella, Seated Statue, p. 249 and n. 12.
13 Parrot, p. 167, pl. XVI a, listed statue P as being in a private collection (read antiquity
dealer) and as Fouilles clandestine de 1924. It was first offered for sale to the Metropolitan
Museum of Art in 1934 by a dealer (Feuardent Frres), in 1937 by another dealer (J.E. Gjou),
and in 1950 by the latters daughter; each time it was rejected. In 1958, the figure surfaced
again, in the possession of yet another dealer (E.S. David), and this time it was purchased
(1959). In 1950, the curator (C.K. Wilkinson) who purchased the statue in 1959 thought that
the head was genuine but that the body was not.
14 Thus, statue P shares most of its inscription with statue I but its name with statue C.
See Edzard, pp. 3940, 5153, 58.

a copy: the inscriptions ductus (writing style) is aberrant, the stone is not
diorite, the thrones feet are flat and not organically rendered, and the fur of
the hat is clumsy. In a chapter in the same volume, the cuneiformist A. Alster
accepted Johansens condemnation. He believed that the minimal change
in the text between statues P and I could easily have been construed by 1930.
Oddly, neither he nor Johansen mentioned the sharing of the statues name
with statue C.15
More or less based on stylistic determinations, some scholars supported
statue Ps authenticity,16 others followed Johansen and Alster, and still oth-
ers were unsure whether it was modern or ancient. G. Colbow thought that
statue P could have been modeled and copied from statue I and the inscrip-
tion copied from both statues I and C. To her, this might indicate that statue
P is a forgery but also that it is wahrscheinlich echt (probably ancient).17
Others were more certain. E. Braun-Holzinger condemned it on the basis of
the inscription as well as style: poor execution of the eyes, a too fleshy nose,
overly modeled cheeks, wrong ears and hat, and incorrectly rendered cloth-
ing; and both H. Steible and C. Suter, concentrating on the coinciding of the
inscriptions, condemned it as a modem work.18
I find it difficult to reject statue P as a forgeryto accept the negative
stylistic observations expressed.19 I perceive the style and execution of the
face, hands, muscled arms, and feet to be in accord with ancient Gudea
sculptures, and the stone indeed is diorite or gabbro, consistent with the
others. If statue P is a forgery, one has to confront the following possible
scenario: a skilled forger in the 1920s had to have had hands-on access to
statue I in the Louvre, obtained accurate measurementsperhaps made a
mold. Furthermore, he or someone else (see note 25, below) had to have
15 For both, see Johansen, pp. 12, 2425, 34, 39, 5758. To me, Johansens attempt at a few
stylistic criticisms of statue P has no manifest verity and seems to be not substantive; he
spends no time on the crucial shared-inscription issue; and he is wrong about the stone (see
below). And Alster ignores the stylistic issues raised.
16 Aside from its purchasers, see Strommenger, p. 682, 1b 6; B. Schlossman, Portraiture
in Mesopotamia in the Late Third and Early Second Millennium B.C., Archiv fr Orient-
forschung 26 (19781979):60, who sees statues P and I as possibly deriving from the same
workshop; Mller, 51; A. Spycket, La statuaire du Proche-Orient ancien (Leiden: 1981), p. 190
and n. 29.
17 G. Colbow, Zur Rundplastik des Gudea von Lagas (Munich: 1987), pp. 8889.
18 Braun-Holzinger, p. 266; Steible, 9697; Suter, p. 36, n. 26, 33: an obvious fake (for the
inscription issue, see below).

19 Muscarella, The Lie Became Great, pp. 173, 214, n. 59. I have no ax to grind either way: if
I thought it a clear-cut forgery, I would readily proclaim it as such loudly and clearly.
copied the inscription accurately and neatly inscribed it on statue P. None

of these elements is impossible but, taken together, are unlikely.
There are at least two possible connoisseurship problems that merit
further investigation of statue P. The position of the eyebrow at the point it
merges into the bridge of the nose is visible in profile; and where they join, a
thick ridge is formed, also quite visible in profile (Fig. 1b). There also seems
to be a heaviness to the body when seen from the rear, as if Gudea is sinking
into his seat. Although not readily evident in photographs, some excavated
Gudea statues, including the Louvre statue I, have a heavy ridge where the
brow joins the nose that is not visible in profile view. In addition, the latters
eyebrows continue as a straight line farther than that of the New York Gudea
before it dips down, and the curve is not visible in the profile view.20 A
thick ridge, however, is clear in the profiles of several unexcavated Gudea
heads, some of which have been condemned as forgeries.21 The heaviness
of statue P is less significant, however, for while two excavated statues have
more natural back lengths and sit convincingly on the seat, three excavated
statuesincluding statue Ihave shortened backs and appear to sink into
their seats.22
The inscription on statue P presents a more troubling problem that can-
not be ignored or easily explained away. Scholarly understanding of the
mind of ancient Near Easterners indicates that it is highly unlikely they
would have given the same name to two separate statues, each of which was
conceived to be a distinct individual. According to J.M. Asher-Greve, The
self is located in the inseparable unity of body and spirit. Because the self can
be re-created and actually be present in an image, Mesopotamian visual rep-
resentations possessed a form or reality which is difficult to comprehend for
modem scholars The body seems to be that of a self, comprising body
and spirit, which can replicate itself in other manifestations such as statues
The spirit, not a replica but a unique entity, can apparently inhabit several
objects simultaneously.23 A name on a particular Gudea statue identifies the
20 See Johansen, pls. 3839, 41, 4246.

21 The Detroit head (below), the Metropolitan Museums Ur-Ningirsu head, and heads in
the Rijksmuseum, the Prado Museum, and the Israel Museum. Johansen, pp. 4041, considers
all these examples to be modern creations or doubtful; the Israel Museum example is the
most obvious forgery (Muscarella, The Lie Became Great, pp. 173174, no. 1); and certainly
problematic are the Chrysler (ibid., no. 2), Leiden, and Prado examples.
22 Statues B, D, F, H, and I. Compare Johansen, pls. 20, 2631, 36, 40.
23 Julia M. Asher-Greve, The Essential Body: Mesopotamian Conceptions of the Gen-
dered Body, Gender and History 9, no. 3 (1997):452.

individual it embodies and distinguishes it from other Gudea statues. Clau-

dia Suter expressed the name issue thus: Gudea could repeat the same deeds
on two statues; but the name is a different matter, a serious thing. It is cru-
cial for the ritual life of the statue. Through the name, the statue receives
life, is changed from a man-made artifact into a cult object that will receive
regular offering. And it is the name that distinguishes this object endowed
with life from others (italics mine).24 No scholar can disregard these two
observations; they remain, perhaps more than stylistic analyses, crucial to
the authenticity (or not) of the Metropolitan Museums seated Gudea.25
The Detroit Institute of Arts possesses a standing Gudea statue with an
inscription, labeled statue M (Fig. 2a and 2b; F82.64). It first surfaced in
1925 in the possession of a dealer and was purchased and then resold to
the Detroit Museum in 1982. From its first appearance, scholars based their
attribution conclusions on the basis of both stylistic determinations and
subjective interpretations; later its inscription entered the discussion. The
statue was first published by V. Scheil, who rightly noted its unique green
stone and the atypical hand position in which the right hand is held loosely
over the left wrist. He interpreted these aberrations as indicating that the
statue represents Gudea as an adolescent (for which no other parallel was
offered).26 G. Contenau called it une grande statuette, a figure of humility,
an original work compared to other Gudea statues, because (after Scheil), he
claimed, it represented Gudea as a youth. Parrot also shared this view. And
H. Frankfort noted the unusual hand gesture and other divergences from the
Gudea canon, which he interpreted as achieving an inner logic consistent
with the stones color, sculpturing, and elegance.27
Despite these positive stylistic as well as subjective assessments,
E. Strommenger expressed doubts about the authenticity of statue M, while
24 Quoted from one of Claudia Suters personal and generous communications with me.
See also Vogel, 67. Vogel does not mention the specific text problem concerning statue P.
25 This signifiesif this Gudea is a forgerythat a Sumerian scholar must have informed
and guided the forger about copying the inscription of statue I up to the point where the
statues name appeared, and then directed him to another statue in the Louvre, statue C, and
the position where that statues name existed!
26 V. Scheil, Une nouvelle statue de Guda, Revue dAssyriologie et Archologie orientale
20, no. 2 (1925):4143. Little of the substantive issues I raise here appears in the entry for this
statue in the Metropolitan Museum catalogue mentioned in note 1, above: E. Peck, Standing
Statue of Gudea, pp. 430431, no. 306.
27 G. Contenau, Manuel d Archeologie orientale , 4 vols. (Paris: 19271947), II, pp. 721722;
Parrot, pp. 165166; H. Frankfort, The Art and Architecture of the Ancient Orient (Baltimore:
1954), pp. 4748. For the otherwise common use of diorite, see above.
Johansen condemned it outright as a forgery.28 Johansen objected to it on

several stylistic grounds: (1) he considered the hand position to be aesthet-
ically indefensible; (2) the garments lower border design does not exist on
any other Gudeas; (3) the eyelids are of different thickness and do not inter-
sect naturally; (4) the diagonally arranged curls on the hat form a central
ring; (5) the statue is small; (6) its stone is paragonite (correct) and not dior-
ite. In a separate chapter in Johansens monograph, B. Alster contradicted
Johansens negative conclusion. Alster defended the statues inscription as
manifestly written by a Sumerian and impossible for a modern scholar to
have forgedhe pointed out that it contains a pun in Sumerianand that
the sculpture must be an ancient work; he thus chose to ignore the stylis-
tic issues raised.29 In a review of the monograph, H. Hrouda understandably
expressed his puzzlement and confusion about the ancient-modern prob-
lem raised stylistically by Johansen and linguistically by Alster. He noted
that the hand-over-hand position did exist in other Gudea representations
but only for deities, and although the inscription evidence was powerful, the
piece might be modern.30
Earlier, in an unpublished manuscript, the Sumerologist D.O. Edzard
expressed doubts about the authenticity of the statue.31 Although primarily
a cuneiformist, he noted the obvious stylistic aberrations: the hand position,
the hats curl deviations, the garment hem decoration, and the statues slen-
derness. He also noted problems regarding the inscription: its proportion to
the statues height; an unexpected cut inscription line that is extended too
far; the cartouche that is too long; the inscription itself, which is not orig-
inal, aside from containing the ex-voto name. If one doubts the statue, he
added, one must assume that a scholars advice had been used in making
the inscription. He himself had strong doubts about the statues authen-
ticity. In addition, in private communications, two other cuneiformists also
raised the possibility of an unscrupulous scholar who could have helped
in the creation of the inscription.32 Alster also addressed this issue but
28 Strommenger, p. 681, 1a [8]; Johansen, pp. 22, 34, 40.

29 Alster, in Johansen, pp. 4956.
30 H. Hrouda, in Zeitschrift fr Assyriologie 64 (1979):151the Stoclet Gudea liesse sich
noch opfern as modern.

31 D.O. Edzard was shown the statue and asked to give his opinion about it. His manuscript
is in the Department of Ancient Near Eastern Art of the Metropolitan Museum of Art and is
dated 19 February 1973. Edzards criticism of the inscription has not hitherto been mentioned.
32 Their opinions were given in communications, dated 3 February 1977, on file in the
Metropolitan Museum of Art. Also mentioned by them is the name of a Sumerologist who
was considered by scholars in Copenhagen to have been the one who aided the forger in
concluded that it was not possible in the context of knowledge of Sumerian

in 1925.33
Although a few other scholars continued to express doubts, the statue
also had its defenders. Spycket, Colbow, Mller, and Hansen have accepted
the authenticity of the statue.34 To Hansen, the hand position was known in
ancient times, the eyes and heavy curving lids were its most prominent
feature, the garment hem decoration is totally believable, the small size
is irrelevant, and the stone has a natural beauty. Hansen further believes
that it is carefully modeled by a fine artist, and, more, it is one of the mas-
terpieces of ancient Near Eastern art. Arniet, Braun-Holzinger, and Steible
also accept it as ancient without argument, while Suter, essentially joining
Johansen and Hrouda, expresses concern, claiming that the authenticity
of this Gudea, along with others (which includes statue P) remains dis-
putable.35
Hansen, the statues most vigorous defender, acknowledges the unique-
ness of the garment decoration and the nature of the curling hat fur. I suggest
that in addition to those aberrations noted by other scholars above, other
deviations remain to be considered: the eyes appear to be improperly exe-
cuted (mentioned by Johansen), are uneven in size and shape, and seem
to look in different directionswhich indeed make them the most promi-
nent feature; the eyebrows are uneven in size and thickness; the ears have
a strange lobe pendant; the mouth pouts and seems too thin; the broken-
away feet area appears to be unformed; and the thumb of the right hand
is unnaturally large and poorly sculpted. I find it difficult to ignore these
stylistic problems, which collectively do not appear to characterize the work
of an ancient craftsman. Consequently, I suggest that it has not been not
demonstrated that statue M is ancient, and thus that a definitive judgment
be withheld for now.
creating the inscription on the Gudea statue in their museum, a statue that they and others
think is a modern creation. See Johansen, pp. 1424, 40.
33 Ibid., pp. 5253.
34 Spycket, p. 190; Colbow, pp. 8586, 8889; Mller, 51; Hansen, 1417. Hansen also men-
tions that there is a curious aspect of the statue, a misalignment of the vertical axis, which
is a difficult trait to explain. And I find it difficult to understandwhether or not the statue
is ancient or modernhow one could claim that the statue is masterpiece: to my eyes, the
piece is grossly sculpted and inferior in aesthetic aspects to Gudea statues, not to mention
other ancient productions. It would be self-serving to label the statue a provincial ancient
work, ancient but uniquely uncanonical: no scholar can demonstrate this for an unexcavated
piece, and it is an old argument used to defend the indefensible.
35 P. Amiet, La Statue d Ur-Ningirsu reconstitue, La Revue du Louvre 24 (1974):243;
Braun-Holzinger, pp, 158, 265; Steible, 84, 100102; Suter, p. 36, n. 26.
Here is a conundrum. Because of its duplicated inscription, the Metropoli-

tan Museums Gudea is legitimately challengedeven condemned as a
forgery by a number of scholarswhile, on the basis of stylistic analysis,
it is both accepted and condemned by scholars. Because of its inscription,
the Detroit Institutes statue is both defended and challenged, and it, too, is
either accepted or condemned on stylistic grounds. On the one hand ; on
the other Both statues present archaeologists with the difficulty of objec-
tively and independently interpreting their inscriptions and their style.
Both features in fact conflict with each other, and the arguments/answers
to resolve this paradox offered by specialists on both sides of the issue
disagreethere is no consensus. I have found it difficult to condemn as
a forgery the Metropolitan Gudea based solely on stylistic grounds; how-
ever, I fully appreciate the negative indications of the duplicated personal
name.36 But with a reversed perception here, on stylistic grounds I have great
difficulty accepting the Detroit Gudea as unquestionably ancient, although
here, too, I recognize the possible importance of its inscription, which some
defend as ancient.37
The Gudea/not Gudea problem is but one of many situations resulting
from what I call gifts of the bazaar, a modern locus that often, by its nature,
denies cultural and chronological certainty and that frequently requires
the expenditure of much time and energy by scholarsas documented
herewho seek honest resolutions to questions. What remains is this:
the consequence of the variety of observations (and subjective opinions)
signifies that both Gudea statues are problematic; therefore, attributions
should be kept abeyant since these works cannot with certainty be said to
be ancient statues (personalities) of Gudea.
36 It would be a rationalization to propose that the statue itself is genuine, perhaps
plundered in 1924, and the inscription added in modern times. All the Tello seated statues
have inscriptions; a seated statue in the British Museum, perhaps of Gudea (see n. 9, above),
bears no inscription (Johansen, p. 14, pls. 5354).
37 See Muscarella, The Lie Became Great, p. 173, and above.
chapter thirty-six
THE VERACITY OF SCIENTIFIC TESTING BY CONSERVATORS*
Solange es Menschen gibt, die Antiken

sammeln, wird es auch Meuschen geben,
die Antiken flschen.
E. Unger 1957: 5
As an archaeologist concerned with studying artifacts made by ancient

cultures I have been involved with confronting and exposing forgeries of
alleged ancient artifacts, antiquities, for decades.1 I consider myself to be
disinterested, which of course does not mean I am always correct in my
conclusions. Though I feel secure that most of my indictments are accurate,
I have on several occasions reversed or modified an opinion. [viz. Muscarella
2001 (2005): 180, 181182, 198, note 4]. But I assert with force that there exists
a large number, thousands, of forgeries of ancient Near Eastern antiquities
in dealers shops, museums, and private collections all over the world; and
they have been made and sold for more than a century. Following an intense
increase in their manufacture after ca. 1950, some scholars have become
more alert to the forgery problem, but ignorance and innocence persists,
and scholars continue to publish forgeries as genuine ancient artifacts. Too
many scholars (and many students) also remain unaware of the magnitude
of the forgery issue: the full extent of its prevalence and the modern cultural
* This chapter originally appeared as The Veracity of Scientific Testing by Conservators,
in OriginalCopyFake?: International Symposium, ed. Stiftung Situation Kunst in Bochum

(Bochum: Philipp von Zabern, 2008), 918.
This paper has been modified from the original lecture format to suit a written paper.
It is essentially the same paper delivered in Bochum, with additions and bibliography.
(The Bochum Conference was one of the most vigorous and open-minded I have ever
attended. Frankness and a willingness to share different, even opposing views, frankly and
as colleagues prevailed throughout: thanks to our hosts, Alexander and Silke von Berswordt,
and their helpful assistant, Hilke Wagner). I wish to thank two conservators and a scholar
for reading this paper and making valuable suggestionsand for correcting and modifying
some scientific misunderstandings on my part.
1 Muscarella 2000, and bibliography on pages: 223224; for use of the words artifact and
antiquity, see p. 17.

932 chapter thirty-six
causes for its existence; and the degree to which forgeries have corrupted
our knowledge of the worlds ancient material and spiritual past.
I submit also that among scholars and the public at large there is another
crucial related problem about which they remain quite uninformed. This
concerns the methods employed to detect forgeries other than the hands-
on, direct, visual investigations anchored in the empirical knowledge of
archaeologists and art historians, an undertaking usually labeled (some-
times pejoratively) connoisseurship. I address here the contentions of
some scientists and conservators that their technical analyses and conse-
quent conclusions are based on objective scientific analysis, and thus are
more accurate than connoisseurship observations, and they are complete,
and (in some instances) final rulings.2 Scholars, students, and the public
are conditioned to believe such claims, that a scholars subjective inves-
tigation must yield to a conservators objective determination, which is
invoked as impartial, regarding whether an alleged antiquity under inves-
tigation is ancient or modern, even if the conservator is a paid employee of
the objects owner (viz. McDonald 2006; Spier 1990: 627; and infra). And it
took me some time to comprehend this perception, to recognize that con-
clusions submitted in scientific or conservator reports can be as incorrect, or
in conflict with conclusions of their colleagues, as are empirical, connois-
seurship determinations by archaeologists and art historians. Further, some
scientists/conservators casually segue from an apparent technical/scientific
analysis into an interpretation or explanation about technical or stylistic
oddities or aberrations, issues not strictly within their purview (viz. Mus-
carella 2000: 738, nos. 40, 41 for examples). The assumption that subjec-
tive stylistic analysis is distinct from objective scientific analysis is a false
dichotomywhich a number of honest conservator/scientists recognize.
The individuals involved in both forms of analyses are subject to fallibility,
ignorance, and erroras well as personal considerations, which includes
fear, intimidation, and dishonesty.
Conservators, like archaeologists and art historians, have no magic
machines; they are not infallible engineers. J. Spier (1990: 623) put it directly
and accurately: Many technical and scientific studies, however, are not con-
clusive, especially in determining authenticity, and often appear to
be invoked by archaeologists as a desperate appeal to the unattainable,
2 Some individuals involved in testing antiquities are primarily scientists, and call them-
selves scientists or scientist-conservators; others, more involved in an history but who also
have some scientific training call themselves objects conservators. I use both forms here.
the veracity of scientific testing by conservators 933
objective result rather than as a proper study. Spier 1990 is one of the few
scholars who raise this significant issue in print.3
In addressing this problem, four interrelated matters must be considered.
First, and culturally significant, the issue of forgeries of ancient artifacts
involves only objects that have not been excavated by archaeologists at a
known site.4 Involved are exclusively objects that have surfaced in bazaars
and in dealers shops (all over the world) or offered as alleged family heir-
looms, all objects that are subsequently sold and exhibited by collectors
and museums. Second, all antiquity dealers and auction houses insist that
they never sell forgeries (but they say others do). And every antiquity pur-
chased by museum curators or collectors is ipso facto claimed by them to
be ancient. Third, the skills and the scholarly and scientific knowledge of
forgers functioning all over the world cannot be overestimated. Nonethe-
less, every owner of a challenged antiquity will shout that no forger could
have made the antiquity they purchased, as it is too complex to be conceived
and then made by a modern artisan. The owner will assert that the antiq-
uity is formally correct (two eyes, a nose, mouth ), and that if there are
anomalies, the object is all the more valuable, a rare and thus (financially)
a precious original, an Unikum, made in a provincial workshop (Muscarella
2000: 18). The reality is that many forgers (and their families) have lifetime
jobs for a good reason: museums and private collections are continuously
acquiring their productions. In addition to being good artisansand per-
tinent to the thrust of the Bochum Conferences discussion, forgers also
monitor scientific reports on age-determination techniques. And they and
their sponsors employ scholars and conservators both for alleged stylistic
and scientific analyses and advice, and sometimes to help with creating
verisimilitude.
And fourth, pertaining to common perceptions of the academic training
and skills of scientists and conservators contracted or employed by private
collectors and museums, two issues must be emphasized. As with archae-
ologist and art historian connoisseurship studies [Muscarella 2001 (2005) is
3 His article contains a balanced discussion of the merits both of traditional connoisseur-
ship and scientific analyses and disagreements and discrepancies between them, recognizing
the problems with both methods. He supports the former more than the latter with some
reserveas I myself have done, and he correctly notes that too much unexamined reliance
has been assumed for the latter; see also pp. 626, 628. His report on a fragmentary forged
torso, alleged through a connoisseurship study, to be close in style to the Getty kouros, is less
satisfactory.
4 For exceptions see Muscarella 2000: 181, no. 24, and page 215, note 61, where I refer to
the salting of objects into sites by excavation workmen, also mentioned by Unger 1957:6.
an example], conservators can honestly err in their conclusions, by igno-

rance, misinterpretation, or miscalculation.5 And not all conservators are
fully trained scientists aware of the full range of scientific techniques. More
important, because it remains little known and involves a sensitive situ-
ation. some conservators are not the disinterested agents that they and
their employers proclaim. Some obey the commands (overt or insinuated)
of their paymasters (as do some archaeologists, curators and art histori-
ans). Some scientists/conservators, when recognizing they are examining
a forgery, will avoid conducting a complete examination, omit tests that
would demonstrate the object is a modern creation, or will ignore or mis-
state obvious problems. Succinctly put, they lie or dissimulate (Brent 2000b:
31), and present definite or nuanced conclusions that point to the ancient
manufacture of the object: because they do not want to offend their employ-
ers. The president of an academic institution defending a large gift of ori-
ental, non-excavated, antiquities that was authenticated by a contracted
conservator (her employee) but challenged by others, revealed it precisely:
Authentication is actually a business procedure (McDonald 2006). This is
a rare claim regarding the process.
That these facts remain unknown to some scholars is illustrated by a
recently recorded automatic deference to a conservators conclusion. A rep-
utable art historian had strongly expressed an opinion in writing that a
bronze said-to-be Roman statue in a private collection was a forgery (a
view shared by others). When subsequently informed by its owner that the
statue had been examined by a laboratory, which report implied that it was
ancient, the scholar immediately retracted her indictment, informing the
owner that she saw no reason that you should not accept the conserva-
tors conclusions.6 Some museum-employed conservators are also reluctant
to inform their director or curators that they have purchased or accepted as
a gift from a donor a forgery. This is indicated by the fact that some museums
that exhibit forgeries as ancient artifacts have a conservation department,
or hire conservators on contract (McDonald 2006); sometimes they publish
their determinations (see also below). I argue that if one has strong nega-
tive feelings about a museum object, one should not automatically accept
the positive conclusions of a museum-employed conservator regarding an
5 A number of participants at the Bochum Conference noted that conservators recognize
this, inferring I was speaking the obvious. But indeed, non-scientists/conservators do not
know this: because few conservators openly write or speak about such matters.
6 Suzan Mazur The Odyssey or Stuart Pivars Roman Bronze Boy, on the Internet:
SCOOP, December 22, 2006: 79; see again Spier 1990: 627.
Fig. 1. 3-Dollar bill.
object his/her museum has purchased. The same situation also exists for
objects donated by a rich donor. I give an example where a museum scien-
tist suffered when disagreeing with a museum curator. This curator believed
that one of his (many curators believe what they curate belongs to them)
antiquities was ancient, but when someone (not a scholar) publicly claimed
it to be a forgery, he publicly accepted the claim. However, when it was sub-
sequently demonstrated by a museum conservator that the object was in
fact ancient, and that the curator had been caught making a mistake in pub-
lic, he actively campaigned to have the conservator fired. The curator then
protected himself by creating a special exhibition for the object, stressing
how and why he always thought it to be genuine.
I turn to a critique of the various technologies that scientists and con-
servators employ to determine the authenticity of various materials by pre-
senting a few examples. I begin with an appropriate model, a 3-dollar United
States bill in my possession (Fig. 1). Everyone, including foreigners, is aware
that no such U.S. denomination exists. But I propose that I could employ
a specialist censervator who would write a report stating that the paper
looks good, the number 3 is in all four corners, there is a presidents face in
the center; and, to explain the oddities, conclude that the bill was surely
printed in a provincial mint. Therefore he sees no reason to doubt its
authenticity. (I am of course being sarcastic, but not formally inaccurate.)
Terracotta: Thermoluminesence (TL) testing is normally conducted by
scientists to determine the age of fired terracotta artifacts, For many years
it has been acclaimed as a definitive scientific technique: thermo-
luminesence testing has become the standard method of testing the
firing dates of ceramic objects. Nearly all important ancient [sic] ceram-
ics are routinely subjected to thermoluminescence testing to remove any
Fig. 2. Hittite Vessel, Cleveland Museum of Art.
doubts regarding their authenticity; and one of the best scientific test-
ing techniques for dating (Meyers 1997: 314); see also Spier (1990: 628)
who accepts TL unhesitatingly. The reality is otherwise: TL is by no means
as accurate as proclaimed. Some scientists and scholars have known for
decades that TL dates are not per se precise, that the technique is very com-
plex; and there is a statistical error of +/- %8.5. Further, at least recently
some forgery laboratories have engineered a system of falsifying TL results,
so that a forgery baked last week can be treated to reflect an age of millennia.
Forgers also make pastiches by bonding pieces of ancient terracotta artifacts
together with modern units, or by adding ground-up ancient terracotta frag-
ments to objects fashioned from locally gathered clay (see also below). TL
testing will thereby authenticate the modern creation as an ancient artifact
(Brent 2001 a, and b). Forgers all over the world know such techniques, as
also do scientist-conservators, some of whom consciously ignore them for
reasons given above. But not all archaeologists know this.
A said-to-be Hittite vessel in the Cleveland Museum of Art (Fig. 2) is a
manifest, embarrassing forgery; it was purchased in 1985, and published
as a small photo in 1986 (Muscarella 2000: 143144, no. 6). I wrote to the
museums curator Arielle Kozloff, who purchased it, requesting that the
vessel be TL tested. She responded briefly on a post card that the vessel
had been TLd before the purchase by Oxford University. After a further
request, she sent me the Oxford report, which stated that the vessel was
made between 1600 and 300bc.
I contacted the Oxford Laboratory, whose scientist-conservator re-
sponded to my questions that the vessel was tested not for the museum
but for a dealer, who had sent in the sample, and that the sample had been
taken from the base. I should say allegedly taken, for the dealer could have
taken a sample from a genuine Hittite vessel, claiming it to have derived
from the one subsequently sold to Cleveland. I was also informed by the
Oxford conservator that unwitnessed testings and sent-in samples were no
longer accepted (compare, however, her statement to the contrary in Brent
2001b: 31), but the sampling from the vessels base was defended: I doubt
very much that the base is genuine and that the rest of the vase fake but
it is of course a possibility (on this wrong assumption relating to another
ancient-modern pastiche, see Brent 2001a: 29). My doubts were thus chal-
lenged. I sent Kozloff a copy of the Oxford report requesting that she check
the base-body construction and take a TL sample from the body; there was
no response. And this neo-Hittite vessel is still exhibited in Cleveland to edu-
cate students and the public about ancient Hittite culturebecause the
museums curator and a scientist-conservator said that it is ancient, than
which there can be no higher authority. I had assumed immediately that the
vessels base was the only part of the vessel TLd: because soon after TL was
developed, forgers began to construct vessels built upon an ancient base.
Every forger is cognizant of the curators code, to never, ever, damage the pre-
cious plundered vessel you have purchasedtake samples for testing only
from the base (some curators are becoming aware of this).
A figurine accorded a Hacilar birthplace by its vendor (Fig. 3) was man-
ufactured to be all the more valuable because it depicts neolithic period
lovemaking, a rarity (none has yet been excavated). It was tested for a Frank-
furt dealer in 1967 by a conservator at the Rmisch-Germanisches Zentral-
museum in Mainz (Muscarella 2000: 139140, 449, no. 58), whose statement
(written in English) is presented as a scientific reportwhich was shown to
potential purchasers by the dealer. The conservator reported that the loving
couple (and another Hacilar forgery made by the same artisan: Muscarella
2000: 139, 450, no. 59) was tested by several methods, some of them chem-
ical. Not one of the tests is described, although one would assume that TL
would have been the primary test. I quote the conservators conclusion: I
Fig. 3. Hacilar Erotic Figure.
do not see any possibility to doubt that both idols are genuine Hacilar works
of art [and] are extremely valuable [and] absolutely uniquea perfect
example of a conservator obeying and serving his patrons wishes. Indeed,
the figurine is unique; it is an authentic modern Hacilar villagers creation
made to be sold, which the conservators paid-for advertisement is intended
to facilitate. These are but two of many scores of Hacilar forgeries housed in
museums and private collections in Europe, the United States, and Turkey
(Muscarella 2000: 135141, 434453).7
A terracotta plaque (Fig. 4) recently offered for sale at Sothebys, New
York,8 is labeled Assyrian, dated, we are informed, on the basis of TL testing
to the 1st half of 1st millennium bc, or earlier. (!) No details are given,
7 Conservators have no monopoly on authenticating Hacilar figurine forgeries. James
Mellaart, the excavator or Hacilar, authenticated forgeries of Hacilar figurines (Muscarella

2000: 140, nos. 60, 61). I also have written and verbal records that Mellaart authenticated as
genuine a number of figurine forgeries that he was shown by dealers and collectors in the
United States.
8 December 6, 2006: no. 97. Also quoted is a report from an alleged scholar of ancient
Near Eastern art.

Fig. 4. Sothebys Assyrian Plaque.
such as who tested it, where, and what specific results were revealed; the
announcement itself suffices. It is to my eyes a badly made forgery, a modern
$ 3 bill forgery (was it radiated?).
A clay plaque (49.5 37cm.) depicting a frontally nude goddess in high
relief was recently purchased (2003) by the British Museum (i.e. by British
taxpayers; Fig. 5). That it was born in the bazaar, and not excavated, is never
mentioned in the museums publications, as such information is considered
to be unimportant by museum staff; it most certainly has not been correctly
identified as coming from ancient Iraq (Collon 2005: 5). First published in
Fig. 5. Queen of the Night, British Museum.
The Illustrated London News (June 13, 1936) it was soon named The Burney
Relief, after its vendor, but following its purchase the museums director bap-
tized it The Queen of the Night (Cellon 2005 11). It has recently (casually)
been assertedsoweit ich seheto be ancient (W. von der Osten-Sacken
2002); also by Cellon (2005). However, from all analysis of all its cultural
and stylistic elements in depth, Pauline Albenda (2005) has argued that the
plaque is most probably a forgery. She discusses the plaques modern his-
tory, its many defenders, and an earlier challenger, Dietrich Opitz (in 1937).
She has painstakingly demonstrated that The Queen is utterly unique in
motif, composition, and style, for example no owls are known in ancient
Fig. 6. Diana, St. Louis Museum.
Near Eastern art. The British Museum curators cite TL testing (from 1975),
which gave a range across 2000 years, 172525bc, as conclusive evidence
manifesting ancient manufacture (such a date range per se is not a nega-
tive). Tests were taken from the rear and front of the plaque: but not on the
ladys body, which may be a crucial omission. I contend that the plaque is not
manifestly ancient, and whether it is a forgery or genuine remains an issue
for scholarly discussion (soweit ich sehe); More refinements have developed
in TL testing, and new testing is therefore mandatory, and also on the body
(for one wonders if the body is a modern addition to an ancient terra cotta
plaque).
In 1953 The City Art Museum of St. Louis, Missouri, purchased a frag-
mentary four-foot high statuette and labeled it as an Etruscan Diana with a
stag (Fig. 6a), and ecstatically proclaimed it to be one of the finest Etruscan
works in existence. When purchased it consisted of many broken pieces,
considered by the purchaser (and intended by the forger) to manifest its
ancient age; subsequently it was assembled by a well-known restorer (see
below, Fig. 6b). Published three times as a masterpiece of Etruscan statuary
by the museums director Perry T. Rathbone (who was a specialist in Amer-

ican art!), it was subsequently defended outrance by him and another
director, also by hundreds of scholars (of American art?), and presumably
the museums conservation department. In 1968 it was TL tested at two lab-
oratories (Oxford and the University of Pennsylvania), where both gave its
age as ca. 40 years old. Rathbone was correct in one respect, it was made in
Italy, in the early 1870s by Alceo Dossena (Muscarella 2000: 24).
Gold: Some scientists who examine alleged ancient gold artifacts claim
that they can detect the percentage of helium particles present, the higher
the amount the earlier the date. But the percentage of helium in gold is
small, and difficult to determine. There is also reference to patina sur-
face analysis and use of a binocular microscope to distinguish ancient
from modern-made gold artifacts (Meyers: 1997: 316). But a conservator
(personal communication) vigorously rejected this claim as insufficient,
subjective, and also noting that it omits a crucial test, that for surface
depletion, which is easy to detect. This lest employs a methodology that
determines surface depletion of the silver and copper trace elements, and
is reliable and I am informed not easy (or possible) to fake. No conser-
vator can conduct an examination on a gold object and declare as sci-
entific fact that it is ancient without explaining the analytical techniques
used.
Ulrike Lw (1998: 346354) and I (Muscarella 2000: 3738, nos. 40, 41) have
argued that based on style and manufacture analysis, several electrum ves-
sels (two are shown in Fig. 7a, and b), purchased in a bazaar, sold as from
western Iran, and curated in the MIHO Museum in Japan are probably
modem forgeries.9 They were apparently made in the same workshop, and
deviate significantly in style and manufacturing techniques from known
excavated examples (from Markik in northwestern Iran). My condemna-
tions were made pace a scientists interpretation that they were genuine
(Meyers in Ancient Art 1996: 177178), and that unparalleled manufactur-
ing techniques indicate that the vessels are even more ancient than known
excavated examples: although this contention lies outside the scope of a
conservators competence (Lws perceptive indictments were made prior
to Meyers 1996 comments).
9 T. Kawami published them as genuine and ancient in Ancient Art 1996: 2631, and Miho
Museum Southwing, Miho Museum 1997: 5152, nos. 10, 11; she mentioned but ignored the
lack of excavated parallels.
Fig. 7a, b. Vessels, NW Iran, MIHO Museum.
Silver: A cast silver plate, once in the Schimmel collection (Fig. 8), can
be considered a forgery on the basis of the animals execution and style
(Muscarella 2000: 60, no. 4, 336). Schimmel sent it for authenticity analy-
sis to the laboratory in Mainz (the same mentioned above regarding the
Hacilar figurine, although a different conservator wrote the same mean-
ingless type of report). When the report was shown to me, I was surprised
(I was innocent then) that there was no mention of inter-granulation or
corrosion, etc. But the conservator claimed ipse dixit: Es gibt keinerlei
[sic] Argumente gegen die Echtheit der Silberschale Notwithstanding
this authoritative, objective conclusion, another scientist examined it
subsequently and presented definitive evidence that the plate is a modern
construction. Two conservators, two analyses, and two utterly different con-
clusions.
Fig. 8. Silver Plate, Schimmel Collection.
Recently surfaced, a silver forgery of a griffin vessel (Fig. 9; not a rhyton

as it has been labeled) is generously equipped with three funnels, one in an
unsuitable place for an ancient object, but appropriate in the modern world.
It is a failed attempt by its modern creator to make it look ancient Iranian
(probably from the well-known Kalmakarra Cave plunder); he did succeed,
however in making it look modern Iranian.10 Purchased by a Metropolitan
Museum of Art Trustee, Paula Cussi, for ca. one million dollars, the griffin
was confiscated by the United States Department of Homeland Security (on
a legal technicality); and the purchase it price it was subsequently returned
by the vendor (H. Aboutaam). I was informed by a reliable source that
three (not one or two) conservators authenticated the vessel as ancient.11
We eagerly look forward to reading these reports. Bronze: Some facts are
10 The object remains unpublished, although photographs are available on the Internet.
11 For a record of one conservator involved see Jean D. Portell, Art Crimes and Conserva-
tion , American Institute of Conservation News, 330: 6: 8.
pertinent: forgers are unable to reproduce difficult-to-detect corrosion and

patina, and inter-granulation testing is mandatory. And data can be manip-
ulated. Some time ago I was given a conservators report prepared for a
collector, about an object that I believe is a forgery, but which the report
claimed was ancient. Because the patina information was vague, I showed
it to another conservator, who read it, laughed, and told me: The conserva-
tor omitted a crucial test which would surely have determined whether or
not the object was ancient. He must have done this consciously, knowing
the piece to be a forgery, because he did not want to upset his employera
common practice. I had sensed (intuited) that something was wrong with
the report, and my colleague confirmed that I was correct.
Also to be acknowledged is that information derived from excavations
can result in important changes in knowledge about ancient metal alloys:
viz. we now know that brassan alloy of copper and zincexisted for
centuries before the Roman period. Since the 1960s we now know that the
Phrygians made brass in the 8th century bc, and earlier excavated examples
have been identified.12 Hitherto, some genuine antiquities were innocently
condemned as forgeries solely because of the presence of zinc.13
A said-to-be Sumerian bronze head (Fig. 10) came to my attention in 1985
via J. Riederer of the Staatliche Museen in Berlin; it has been labeled ancient
by two scholars (Strommenger 1994: 125, note 1). I replied that I believed it
to be a forgery; Riederer agreed. In 1987 a conservator from another Berlin
museum (Museum fr Vor- und Frhgeschichte), whose curator had pur-
chased it, declared in an inadequate report that it was ancient (Born, 1987).
Soon after, E. Strommonger, the wife of the museums curator, published
it as ancient (Strommenger 1991).14 In the same year, Riederer published a
report explaining clearly why it is a manifest forgery (see also Riederer 1994:
262265); although this report was rejected by Strommenger (1994: 125), cit-
ing scientific analysis (by H. Born?). In Muscarella 2000 (162, no. 22, 474) I
declared it a forgery. Thus, we have two conservators and four scholars dis-
agreeing on its age, in a classic science vs. connoisseurship confrontation.
All are equally fallible.
12 A. Steinberg et al., in Rodney S. Young, Three Great Early Tumuli, Philadelphia. 1981:
286288.
13 Compare the conclusions of J. Riederer, Die Erkennung von Flschungen kunst- und
kulturgeschichtlicher Objekte aus Kupfer, Bronze und Messing durch naturwissenschaftliche

Untersuchungen, in Berliner Beitrge zur Archometie, 2, 1977, 8595.
14 Here repeating a previous defense of two forgeries purchased by her husband: Mus-
carella 2000: 136137. 438, 439, nos. 13, 14.

Fig. 9. Griffin Vessel.
Fig. 10a, b. Berlin Bronze Head.

In 1974, a photo of a stone head was sent to me by a dealer in Beirut

(Fig. 11, left; Muscarella 2000: 161, 470, no. 14). I suggest it is a badly made
forgerya copy of the head of Puzur-Ishtar; his torso was excavated in
Babylon, but the head was later recovered from a dealer (Fig. 11, right). In
1985, Riederer sent me a photo of a bronze head excavated in a Berlin bazaar,
and which I recognized to be an exact copy of the Beirut stone head (Fig. 12,
left; Muscarella 2000: 161, 471, no. 16). In 1988, Riederer sent me a photo of
another bronze head, a copy of the 1985 bronze head, also from a German
dealer (Fig. 12, right; Muscarella 2000; 161, 471, no. 15). One bronze head is
in Berlin, the other in Stuttgart. Riederers 1991 publication demonstrates
that the two bearded heads are modern creations. In 2000 I published both
heads as forgeries (Muscarella 2000: 161, nos. 15, 16, 471). Thus again, at least
two conservators (possibly more) and several scholars disagree with these
example regarding the issue of ancient vs. forgery.
Stone: Here there is much controversy about age determination tech-
niques, which are difficult to understand for a non-scientist. If cleaning
is recognized, when did it occur? Can we determine if it resulted from
chemical or mechanical action? We often hear this weathering must have
taken many centuries to develop or such weathering or the presence of
lichen could not have been developed in a laboratory. Spier (1990: 630) per-
ceptively wondered (regarding the Getty Museum stone kouros) whether
forgers could successfully fake de-dolomitising, which activity was in fact
demonstrated to exist shortly thereafter (New York Times August 4, 1991:
25; Riederer 1994: 274; Lapatin 2000: 47). Concerning the presence of cal-
cium oxalate, considered significant for age determination, does it occur
only from natural aging, or can it be produced by modern immersions in
oxalic acid (Lapatin 2000: 50)? From discussions with conservators, how-
ever, it seems that oxalic acid has no value for age determinations for marble
and limestone. One cannot accept any claims that cite the above as man-
ifest evidence of old age for a stone artifact.15 The Getty kouros is a good
example of the problem. (Lapatin 2000 is a good summary of the purchase
process and conservator attempts at analysis and authentication involving
this kouros).
15 As for example, the eager acceptance by W.G. Lambert concerning methods employed
at the. University of Georgia, U.S.A. that allegedly authenticate alleged ancient stone artifacts
(N.A.B.U, 2004, 3: 61).
Fig. 11a. Berlin Stonehead. Fig. 11b. Puzur-Ishta Head.
Fig. 12. Bronze Heads.

Wood: Dendrochronological examination of wood remains is universally

employed to date archaeological sites, and is considered to be one of the
major chronological tools. In recent years, however, it has been extensively
researched, which resulted in vigorous criticisms about matters of which
archaeologists have been unaware. The critics have raised essential tech-
nical and scientific issues that must be comprehended before one accepts
dendrochronology-derived dates: the cross-dating of trees across conti-
nents, the use of different tree species for ring comparisons, the lack of
bark on a specimen and its consequences for accurate dating, and further,
awareness that the analysis functions on a floating chronology. There is
an additional and equally significant problem with dendrochronological
dating, namely the demonstrated re-use of wood in ancient times, which
economical and preservative activity is still practiced in the modern Near
East, viz. reuse of all wooden parts such as wooden beams, frames, etc.,
that are removed from a structure after an earthquake and transferred to
a newly built structure. A perfect example of reuse of wood in antiquity
is documented at Gordion, where in one Destruction Level building (CC3)
three wood beams were recovered that had been cut 400 years earlier than
another sample here;16 and a beam that had been cut 6200 years ago was
recently discovered in a Black Sea Turkish modern-day house was (Keenan
2002: 232; idem 2006: 1112). All of which adds up to the reality that archaeol-
ogists definitely do not have a straightforward system, one providing unam-
biguous date determinations (see viz. Keenan 2002: 232233, idem 2006;
James 2002: 18; Mielke 2006). Dendrochronological dating seems not to be
usually employed with artifacts, but when such analysis is used, because of
the above reasons it could not per se date an excavated or an unexcavated
artifact.17
16 P. Kuniholm Dendrochronology and Radiocarbon Dates for Gordion and other Phry-
gian Sites, Source VII, 3/4, 1988: 6, 8. Unfortunately for archaeological accuracy. Kuniholm
and his Gordion colleagues have ignored this crucial reuse documentation in their later con-
clusions about Gordion chronology: by using the recovered beams in the destruction level as
well as one in the Early Phrygian fortification wall as evidence. Keenan and James present
a strong critique (Mielke less so) of Kuniholms analyses and conclusions regarding the dat-
ing of sites from recovered wood. Note also Keenan 2006 (1316) for a rejection or the recent
dendrochronological dates announced for Assiros in Macedonia by Kuniholm et al. These
dates demand a re-dating of the Proto-Geometric period to the early 1st millennium bc, a
major upward chronological adjustment for that period. Sec also J. Yakars acceptance of the
Assiros dendrochronological analysis and its early chronology proclaimed in the same vol-
ume as Mielke 2006 (119).
17 See for example its use in determining the age of European furniture in The Getty, A
Renaissance Cabinet Rediscovered posted on the Internet. I of course have no opinion on

this item (I owe the reference to Mark Rasmussen).
Fig. 13a, b. Breastplate, Land of the Sible Museum.
Fig. 14. Gold Copy of the Seven Brothers animal-headed vessel.

A note on restorers: restorers normally work for dealers and collectors,

some believing they are reinstating to its original form an ancient artifact.
Forgers, however, often damage their creations to give them a sense of age
and verisimilitude, and then employ restorers as launderers to restore the
modern-made breakages in modern-made artifacts. Some restorers do this
knowingly, as normal business, but I knew one utterly honest but innocent
man, Joseph Ternbach (Muscarella 2000: 24), who did not know he was given
forgeries to restore by dealers and collectors. Among the forgeries he was
deceived into restoring were the St. Louis Museum Diana (Fig. 6); the Land
of the Bible Museum bronze breastplate (Fig. 13; Muscarella 2000: 181, 499,
no. 25), and a gold copy of the Seven Brothers animal-headed vessel (Fig. 14;
Muscarella 2000: 54, 319 no. 21).
Bibliography
Albenda, Pauline 2005: The Queen of the Night PlaqueA Revisit, Journal of the
American Oriental Society 125.2: 171190.
Ancient Art from the Shumei Family 1996, Metropolitan Museum of Art (abbreviated
here as Ancient Art).
Born, H. 1987: Die Bedeutung antiker Herstellungstechniken zur Beurteilung fl-
schungsverdchtiger Bronzen, Deutsche Gesellschalt fr zerstrungsfreie Prfung
13: 146155.
Brent, Michael 2001a: Faking in African Art, Archaeology, January/February: 2730
2001b: The Limits of TL, Archaeology, January/February: 31.
Colon, Dominique 2005: The Queen of the Night, The British Museum Press.
James, Peter 2002: The Dendrochronology Debate, Minerva July/August: 18.
Keenan, Douglas J. 2002: Why Early-Historical Radiocarbon Dates Downwind From
the Mediterranean Are Too Early, Radiocarbon 44, 1: 225237, Anatolian
Treering Studies Are Untrustworthy, available at http://www.informath.org/
ATSUO4a.pdf.
Lapatin, Kenneth 2000: Proof: The Case of the Getty Kouros, Source XX, 1: 4353.
Lw, Ulrike 1998: Figrliche verzierte Metallgefae, Ugarit-Verlag, Mnster.
McDonald, Robert 2006: Authenticity of Tseng Artifacts Questioned, Daily Sundial,
Internet, May 4, 2006.
Meyers, Pieter 1997: Acquisition, Technical Study and Examination, in Miho
Museum, South Wing, Miho Museum: 313320.
Mielke, Dirk Paul 2006: Dendrochronologie und hethitische Archologieeinige
kritische Anmerkungen, BYZANZ 4: 7794.
Muscarella, Oscar White 2000: The Lie Became Great, Styx, Gronigen, 2001 (pub-
lished in 2005): Jiroft and Jiroft-Aratta, Bulletin of the Asia Institute, 15: 173
198.
Osten-Sacken, E. von der 2002: berlegungen zur Gttin auf dem Burney-relief,
in Sex and Gender in the Ancient Near East, eds. S. Parpola and R.M. Whiting,
Helsinki: 479487.
Riederer, Josef 1991: The Scientific Examination of Forgeries of Mesopotamian

Bronze Heads, in Dcouverte du Mtal, Picard: 439447: 439446 1994: Echt und
falsch, Springer-Verlag.
Spier, Jeffrey 1990: Blinded with Science: the Abuse of Science in the Detection of
False Antiquities, The Burlington Magazine, September: 623631.
Strommenger, Eva 1991: Ein frhdynastischer Statuettenkopf aus Bleibronze im
Museum fr Vor- und Frhgeschichte Berlin, Acta Praehistorica et Archaeologica
23: 141145 1994: Ein frhdynasticher Statuettenkopf aus Bleibronze im Museum
fr Vor- und Frhgeschichte, Handwerk und Technologie im Alten Orient, ed. Ralf-
B. Wartke, von Zabern: 125.
Unger, E. 1957: Flschungen in RLA, Dritter Band, 1957: 59.
Section Five
Forgeries of Archaeological Provenience

chapter thirty-seven
ZIWIYE AND ZIWIYE: THE FORGERY OF A PROVENIENCE*
Abstract: For almost 30 years many hundreds of objects in gold, silver, bronze, and
terracotta have been accepted and published by scores of scholars as having derived
from the ancient site of Ziwiye in NW Iran. All of these objects in fact came from
antiquity dealers, none was excavated by archaeologists. The Ziwiye problem is
discussed from an archaeological perspective that examines the background for
this acceptance. It is argued that a study of the publications of the alleged find
results in the conclusion that there are no objective sources of information that
any of the attributed objects actually were found at Ziwiye, although it is probable
that some were; that dealers and uncritical scholars together are responsible for
treating the objects as an actual archaeological find from one find-spot; and that the
objects have no historical and archaeological value as a group. Further, it is argued
that the methodology employed by many scholars concerned with the material has
been defective and improper and that many art historians and archaeologists have
generally tended to ignore the serious implications involved in seeking firm cultural
and archaeological conclusions when working with objects claimed to derive from
sites that were not excavated by archaeologists.
The study does not attempt to offer its own conclusions regarding the alleged
nature, chronology, and artistic elements of the objects allegedly derived from
Ziwiye; it is only concerned with an analysis of the scholarly treatment of a mass
of material offered for sale by dealers and its subsequent integration into a putative
reconstruction of the past.
1. The Objects and Their Acquisition
The first published reference to the alleged discovery of the Ziwiye Trea-
sure was made by Andr Godard and appeared in the Muse Cernuschi
Catalogue of the exhibition of Iranian art in 1948 (Godard 1948, 914). The
name of the site was not mentioned but was referred to as une montagne
voisine de Sakkiz. Godard cited in passing a gold gorget (pectoral), and
he cited in the catalogue, without photographs, three objects trouvs aussi
dans la region de Sakkiz. Number 5 in the catalogue was an incised red ter-
racotta vessel with a ducks head protome that served as a spout; No. 7 was a
* This article originally appeared as Ziwiye and Ziwiye: The Forgery of a Provenience,
Journal of Field Archaeology 4, no.2 (1977): 19 7219.

956 chapter thirty-seven
pair of silver bracelets, one with lion, the other with ram-head terminals; and
No. 8 was a greyware terracotta ram-header vessel. The first two objects were
listed without explanation as belonging to the private collection of Yedda
and Andr Godard, the third to the antiquities dealer, Rabenou. Neither
Godard nor anyone else ever mentioned that Godard collected antiquities;1
nor did Godard state in this 1948 workor elsewherethat Rabenou was
both the dealer most intimately involved with the sale of objects said to
come from Ziwiye and the commercial dealer working at the site itself (this
was not mentioned in print until Dyson 1963, 34). The two silver bracelets
and the ducks head spouted vessel were mentioned again in Busagli 1956,
Nos. 204 (pl. XXVI) and 208, as still in the Godard Collection;2 and the ram-
headed vessel seems to be the example published in Godard 1950, fig. 57, but
there said to be in the Teheran Museum.3
The first scholarly report on the find was made by Godard before the
Acadmie des Inscriptions et Belles Lettres, which was published as a com-
munication in their 1949 report (Godard 1949). In this report Godard gave
no specific details regarding the find except to point out that it was made by
local villagers who divided the spoils, which were subsequently dispersed
(presumably to dealers). He mentioned certain objects said to have been
found in a bronze container by the villagers but gave no illustrations, and
he presented his opinions about the artistic attributions of the objects, their
chronology, and importance. This information was repeated in Godard 1950
and need not be further discussed here.
On April 8, 1950, in the French magazine France Illustration, Yedda
Godard published for the first time the alleged account of how the find was
made:
1 Thus Godard 1950, fig. 49 is obviously the same vessel as Godard 1948, No. 6, although
no mention is made of this fact: see here also note 8. Dussaud 1951, 287288, stated that
Godard got the Teheran Museum to purchase l ensemble; see also Samedi 1960, 18; Bussagli
1956, 119 implies that Teheran recovered all the finds (even though he includes two objects
from Godards private collection!); Ghirshman 1950, 181 stated that the Teheran Museum
acquired les pices capitales, white Ghirshrnan 1961, 81, and 1964b, 6, claim the major part
was acquired by Teheran. Note also the less than candid comment of Y. Godard 1950b, 714
(quoted in the text). Furthermore, Wilkinsons statement (1975, 7) that Godard did his best
to conserve the treasure in the country of its origin, does not seem to be supported by the
facts.
The need to refer repeatedly to publications in the body of the text in order to document
most conveniently the history of the Ziwiye affair required a deviation from the usual JFA
format for references. A full bibliography is appended at the end of the article.
2 Ghirshman 1964a, 113, mentions silver bracelets but does not give their location; Ghirsh-
man 1973, No. 110, refers to a silver bracelet with a rams head terminal in the Teheran Museum.
3 It is of course possible that the vessel passed from Rabenou to the Museum.
ziwiye and ziwiye: the forgery of a provenience 957
II y a pres de trois ans, une jeune berger qui gardait ses moutons sur une
haute colline du Kurdistan vit briller un fragment dor parmi des dbris de
constructions et de fortifications anciennes. Il se ramassa et pensa quil prove-
nait dune fissure apparue la suite dun prcdent orage. Il appela un de
ses camarades occup un peu plus haut dterrer de ces racines dastragale
qui fournaissent le gomme adragante, et tout deux se mirent largir la
crevasse. Ils en sortirent quelques objects qui leur parurent tre un don de
Dieu. Craignant pour leur nouvelle fortune, les deux enfants lenvelopprent
dans quelques guenilles et songeaient la cacher de nouveau lorsque survint
un Isralite qui, de son ct, etait venu sur la colline pour y recueiller une
certaine terre propice au nettoyage rituel de ses casseroles. Surpris de la trou-
vaille, il offrit de la troquer contre quelques pices de monnaie, ce qui fut
accept.
La nuit suivante le Juif revint sur le site de la dcouverte, trouva dautes objets
dor et les enfouit dans le jardin de sa maison, mais, les enfants ayant racont
laventure, laffaire sbruita, et lIsralite fut menac, molest, pill. Les habi-
tants du village voisin se rendirent sur le lieu de la fouille et creusrent, lun
surveillant l autre. Dautres objets dor furent exhums, et dauprs discus-
sions eurent lieu au sujet de ce qui en reviendrait chacun, la suite de quoi
tout fut coup en morceaux et partag. (Y. Godard 1950a, 331).
Almost three years ago, a young shepherd who was keeping his sheep on a
high hill in Kurdistan noticed a gold fragment glittering among the debris
of ancient buildings and fortifications. He picked it up and thought that it
came from a cleft in the rock that had appeared as a result of an earlier storm.
He called over one of his comrades, who was occupied a little further up
in digging around the roots of an astragalus tree, which supplies tragacanth
gum, and the two devoted themselves to enlarging the crack. They brought
out several objectsto them, a gift from God. Apprehensive about their new
fortune, the two boys wrapped it up in some rags and thought of hiding it
again, when there came upon them an Israelite, who, for his part, had come
up the hill to gather some variety of mould necessary for a ritual cleaning of his
saucepans. Surprised at the find, he offered some money, which was accepted
in exchange for several pieces.
The following night the Jew returned to the site of the discovery, found some
other gold objects, and buried them in the garden of his house; but, since
the boys had recounted their adventure, rumor of the affair spread, and the
Israelite was threatened, harassed, and robbed. The people of the neighboring
village went to the place of the find and dug, each keeping an eye on the
others. Other gold objects were dug up, and, after some discussion about the
allotment of the finds, everything was cut into pieces and distributed.
(Translated by Lesley K. Cafarelli.)
She gave no source for this story, presented as historical facta leitmotif
that characterized almost all future discussions of the Ziwiye treasure. Nor
did she explain why the Juif/Isralite was so designated, inasmuch as he
presumably was one of the local villagers; perhaps the designation was
added to document the intrigue involved. What is more, she hinted at
incidents of intrigue, jealousy, and revenge, but refrained from discussing
them for fear that Les lecteurs croiraient que j ai entrepris d crire un
roman policier.
The same story was retold in English in a slightly abridged version in the
May 6, 1950 issue of the ILN. Here an additional sentence was added, one
of interest: this priceless find was broken up into small pieces and shared
among all the contestants; and has only recently been reassembled
though whether in full is doubtful (Y. Godard 1950b, 714).
For reasons never offered by their authors, the details of Y. Godards story
were metamorphosed in subsequent retellings. Thus van Ufford (1962, 25)
claimed that des laboureurs ont trouv the treasure; Samedi (1960, 17)
said it was found by trois jeunes bergers; and Kantor (1957, 10) claimed,
although not sure it was true, that some women needed fresh earth with
which to repave the floors of their houses. When they dug into the slope of
the nearby mound (sic) , they came upon a bronze sarcophagus ; it was
crammed with objects of precious metal and ivory. Although the women
hastily concealed their find and their men-folk furtively carried it into their
houses at night, its existence could not be kept secret. Y. Godards petit
berger out with his sheep becomes trois jeunes bergers, des laboureurs,
and some women, and the Jew disappears. Even the alleged date of the find,
stated by Y. and A. Godard to be 1947, was not universally accepted: Ghirsh-
man 1950 said it occurred pendant la deniere grande guerre (later, however,
1964a, 98, hegave the 1947 date), followed by Schefold (1954, 428); Sulimirski
(1954, 298) gave the date as 1946; and Carter (1957, 107, 112) casually claimed
that it had been made in 1928. Moreover, the alleged specific location of the
find-spot moved about in subsequent publications. In Godard 1949 (168) the
treasure was discovered au sommet de cette colline; in Godard 1950 (78)
it was changed to quelque distance du sommet de la colline; according
to Ghirshman 1950 (181) it was found dans le flanc d une colline; Porada
(1965, 123) said it was found below the walls of a citadel; and Dyson 1963
(34) reported that the villagers claimed it was found about halfway down
the slope in a deep gully.
The first time the scholarly world saw in print any of the objects said
to come from the treasure was in the two Y. Godard publications of 1950.
She published only four gold objects, but let it be known that there were
numerous fragments. Later in the same year the first major description of
the objects was published (Godard 1950) and it is this volume that is usu-
ally cited by scholars as the starting point and as the standard work on the
treasure. Godard 1948 and 1949 and Y. Godard 1950a and b are rarely cited
and appear to be generally unknown. Paradoxically, Godard 1950 did not
mention the details of the find discussed by Y. Godard; we are given only
a brief undocumented statement that the objects were found in 1947 in a
bronze container, that they were cut into pieces by the villagers and sub-
sequently dispersed (Godard 1950, 78), information already published in
his 1949 report.4 In addition to this information, Godard made a distinction
between objects he claimed derived from the treasure, from the receptacle,
and those objects he claimed derived from the site at large, provenant de
Ziwiy, sur la colline de Ziwiy mais ne font pas partie du trsor, des envi-
rons de Ziwiy, and so forth (Godard 1950, 55, 56, 65, figs. 46, 47, 49, 50, 51,
53, 57, 58; see also Busagli 1956, 122). What evidence, if any, that was avail-
able to Godard for these distinctions was not vouchsafed to us. Moreover, in
a number of cases he did not state the modern provenience of the objects
published.
One receives from this publication the implicit, if not explicit, impression
that the corpus of the Ziwiye materialgold, silver, bronze, ivory, stone,
and terracottais being presented. In some instances objects are simply
cited without description and photographs: on page 7 nornbreuses armes,
poignards , pointes de flches et ttes de lances en fer, des fragments de
siges luxueux, en bois de cypres recouvert de bronze, des bases de toutes
formes are mentioned; and on page 9 a gold vase, a gold cordelire, gold
earrings, necklaces, and pins are cited, none of which objects is further dis-
cussed (one of the gold necklaces might be Dimand 1950, 145, left). Other
objects, however, are more fully discussed and are illustrated. Godard pub-
lished the following 14 gold objects:5 a complete gorget (figs. 1024, 33; also
published by Y. Godard); fragments of a trapezoidal plaque with three dec-
orated registers extant (fig. 25; for the lower part see Dimand 1950, 145); two
joining fragments of a plaque decorated with a hero fighting a (missing)
lion (fig. 27); a fragment of a belt (?) decorated with Scythian style animals
and birds heads (fig. 29); two griffin heads (fig. 30) and two lion heads,
4 The omission by Godard is all the more striking because Y. Godards two. articles
appeared on April 18 and May 6, 1950, while the manuscript of A. Godards book was com-
pleted months earlier, in January of that year (Godard 1950, 126).
5 After some thought I decided to list in the text, rather than place in footnotes, all the
objects with their descriptions and references published by Godard and other scholars (but
see note 13). This choice was made so that the reader could immediately recognize what
object was being discussed and so that he would receive the full impact of the number of
objects introduced over the years.
probably protomes (one griffin was published by Y. Godard); an ibex bracte-

ate (fig. 39); a gold bracelet with lion-head terminals (figs. 4042; only
one mentioned; Y. Godard published one head: in 1950a, said to be from a
bracelet, in 1950b, said to be from a thong);6 a fragment from a scabbard and
decorated with saiga heads en face (fig. 44); a torque (fig. 45); a fragment
of a plaque decorated with recumbent ibexes and stags and en face lion
heads (fig. 48; also published by Y. Godard), and a band with inlaid rosettes
(fig. 90). Note that there was no mention that other fragments of some of
these objects may have existed albeit not published by Godard.
The silver objects consisted of apparent horse trappings: decorated pen-
dentifs (figs. 96100), decorated discs (figs. 101106); belt ornaments (figs.
107108); and a plaque decorated with a rampant lion (fig. 109); a silver vase
is also mentioned and published as a drawing (figs. 53, 54). Of some inter-
est is the fact that the two silver bracelets mentioned in Godard 1948, No. 7,
are not cited here. Godard published a plain bronze pendentif and some
decorated bronze discs (figs. 9395), but made it clear that there are more
examples in bronze and iron than those published. Also published as from
Ziwiye, but not from the treasure, was a fragment of a horse bit terminating
in a Scythian style birds head, and a similarly decorated stone seal (figs. 46
47).
According to Godard (1950, 78) the ivories were indeed part of the con-
tents of the bronze container but were not recovered by the plunderers until
they began to sift the earth sometime after the initial discovery was made.
He published 21 relief fragments, one statuette, and one champlev frag-
ment with inlays (figs. 6689, 9192); no wooden objects were mentioned.
There is an implication that other ivories were recovered, but that most of
these were crushed or damaged by the time their value was realized. Fur-
thermore, it seems that every ivory published was in the possession of one
or more dealers.
The few examples of terracotta that were published were said to come
from the site but not from the treasure: the ducks head spouted vessel
first published in Godard 1948, No. 6 (fig. 49, but with no reference to its
present provenience); a similar vessel, painted (fig. 50); a fragment of a lions
head (fig. 51); one decorated glazed vessel (figs. 55, 56; on page 7 others are
6 The same bracelet was subsequently published many times, but that of Anonymous
1956, 107, lower right corner, is the most constructive for illuminating how the provenience of
unexcavated objects changes in publication: the bracelet is here said to come from Hamadan
and it is conveniently dated to the Achaemenid period.
mentioned); and several animal-headed situlae (Godard calls them rhytons,

but they have handles), two of which are illustrated (figs. 57, 58).
In the same year that the Godards articles and book appeared, Ghirsh-
man published an article in Artibus Asiae about the Ziwiye treasure, called
Le Trsor de Sakkez. Ghirshman claimed that the find was made during
the last war (without giving his source), but gave no details other than that
it was made by the local villagers. He did not mention Godard 1948 and 1949
(Ghirshmans paper may have been in press for a year or more), and he sim-
ply presented it as a matter of fact that the objects were found at Ziwiye.
Only a few of the objects published were the same as those published by
Godard: the gold bracelet with lion-head terminals (as with Godard, only
one was mentioned), the lower part of the trapezoidal plaque, the gorget,
and one lion and one griffin head protome. Moreover, Ghirshman published
additional fragments of objects previously published by Godard: the scab-
bard with saiga heads, the plaque with ibex and stags and en face lion heads,
and a fragment of the cuve de bronze, which Ghirshman stated was a bowl or
basin fragment (see also Ghirshman 1954, 107, and Arnandry 1966, 111, note 1
and fig. 1). Also published with a photograph was a gold earring (fig. 23), a
type of object only cited by Godard. The remainder of the objects presented
as deriving from Ziwiye had not been previously mentioned by Godard: a
decorated bronze plaque (fig. 4); a gold chape (fig. 7); a gold bracelet deco-
rated with duck head terminals (fig. 17); two decorated gold belts (figs. 18
20); a bronze strip (fig. 22); silver pins (fig. 23; Godard 1950, 9 mentioned only
gold pins); gold fibulae (fig. 23); a silver plate with gold inlays (figs. 9, 10); and,
finally, he cited one ivory plaque peut-tre pourrait-on attribuer au mme
groupe la seule plaque en ivoire sculpt que je connaise (183, fig. 5; italics
mine).
Ghirshman concluded his catalogue by giving the impression (as Godard
had previously) that he published the corpus of the alleged treasure: cet
ensemble de quelque vingt objets provenant d un seul et mme trsor
(Ghirshman 1950, 198). Aside from the fact that he did not refer to the other
objects published by Godard, of interest is the fact that only one ivory was
claimed as deriving from the treasure, and one not stylistically related to
those already published.
Godard 1951 was a sharp retort to Ghirshman 1950. Warning that antiquity
dealers arbitrarily assign site labels to objects in their possession, Godard
stated that no less than six of the objects published by Ghirshman (figs. 5;
1720, 22, 23) did not derive from Ziwiye, but from other areas in Iran, and
that the ivory was actually of the Islamic period (Godard 1951, 240241). He
also claimed that the decorated bronze plaque did come from Ziwiye but
not, as Ghirshman claimed, from the treasure; his source was le fouilleur
lui-mme (of which more later). Godard did not discuss why he himself had
not mentioned this plaque in his earlier publication; moreover, he tacitly
accepted as deriving from Ziwiye both the silver plate with gold inlays (qui
semble bien provenir de Ziwiy, peut-tre mme du Trsor, Godard 1951, 242;
italics mine), and the fibulae, but again not mentioning that he had not
originally published or cited them.
The first publication of material said to derive from Ziwiye by a scholar
other than Godard or Ghirshman appeared in 1952. Wilkinson published
a number of ivories acquired by the Metropolitan Museum of Art, all of
which are of the same type of style as those published by Godard 1950. Some
are obviously part of the same panels, and one was the very same ivory as
published by Godard (Wilkson 1952, fig. on page 234; Godard 1950, fig. 85).
Wilkinson (1952, 239) stated that they were found in a bronze receptacle
that was probably a coffin of Assyrian workmanship, obviously following
Godard. He also published a new fragment of the bronze receptacle that had
been acquired by the museum and that was clearly related to the fragments
published by Godard. Wilkinson offered no discussions or opinions about
the nature of the discovery or other objects that allegedly may have been
found at Ziwiye.
A second article by Wilkinson in 1955 published more objects acquired
through purchase or loan by the Metropolitan Museum, from Zawiyeh
many of which were found in a bronze bath or coffin made for an Assyrian
prince (Wilkinson 1955, 213; see also ILN April 16, 1955, 699). These included
the museums large fragment of a trapezoidal plaque, similar to the example
published by Godard. This article informed scholars for the first time that
there were in existence fragments of at least two similar plaques, that there
was another fragment of the ibex, stag, en face lion plaque, and that the gold
bracelet in the A.B. Martin Collection was the mate to the one published
by Godard (where only one had been mentioned; Samedi 1960, 18, 24, still
referred to one bracelet). Mentioned without illustration (220) was a gold
fragment in the Pomerance Collection that was part of Godard 1950, fig. 27.
In addition to these fragments Wilkinson also published as deriving from
Ziwiye a gold plaque acquired by the Metropolitan Museum in two parts
and decorated with the bodies of two rampant lions with one face placed
below a tree (figure on page 216), a piece never mentioned before as part
of the treasure (for a further discussion of this piece see below, Section
8).
The following year Wilkinson published two large silver ram-headed ves-
sels, one purchased by the University Museum in Philadelphia, the other
by the Metropolitan Museum. Wilkinson cited both Ziwiye and the appar-
ently clandestinely dug site of Kaplantu, three miles SE of Ziwiye (cf. Godard
1950, 67, fig. 1; below, Section 7): That the two silver vessels came from the
area of these villages is practically certain (Wilkinson 1956, 9). No docu-
mentation for this strongly worded assertion was offered, nor was it noted
that neither Godard nor Ghirshman had mentioned such objects in their
reports (see also Wilkinson 1967; Shepherd 1966; Ghirshman 1961). That this
type of strong but undocumented certification has itself generated other
equally strong statements with the same thrust is neatly illustrated by the
comments in Mallowan and Herrmann (1974, 55): We may also feel some
assurance in accepting as genuinely belonging to the collection a number
of objects which reached the Metropolitan Museum, New York, during or
immediately after the early stages of licensed digging.
In 1960 and 1963 two more articles on material said to come from Ziwiye
were published by Wilkinson. Wilkinsons 1960 article was primarily con-
cerned with dating the bronze container in which the treasure was allegedly
found, but a few more objects were also introduced (pl. XXX): an ivory from a
private collection (whose Ziwiye provenience was challenged by Mallowan
and Herrmann 1974, 56); a gold fibula with two lions on the arc, and a silver
example (see also Muscarella 1965, pl. 58, fig. 3; the fibula cited by Wilkin-
son in note 10 is gold, not silver as stated; and note that Ghirshman 1964a,
100 mentions 4 gold and 38 silver fibulae from the treasure); and a silver
pomegranate-head pin. Wilkinson 1963 published for the first time two more
fragments from the Metropolitan Museums trapezoidal plaque; a decorated
bronze bucket reputedly found in the Ziwiyeh area (Wilkinson 1963, figs. 14,
15); and another ivory fragment (fig. 16).
The most recent report by Wilkinson on Ziwiye was his 1975 publication
of the ivories and miscellaneous objects in the Abegg Collection. Thirty-
seven ivory fragments exist in the collection, twenty-two of which conform
in style to those published by Godard in 1950. There was no hesitation
evident in assigning not only these fragments to the treasure, but the others
as well. Thus, although Catalogue No. 21 was singled out as not belonging
by style to the group hitherto described, it is claimed for the treasure: In
spite of this [the differences], there seems no reason to believe that it belongs
anywhere else than in the Ziwiye treasure (italics mine; Wilkinson 1975, 53).
That statement implies that there is some reason, albeit not revealed, for
assigning it to the treasure. Likewise, the ivories nos. 2437, all examples
hitherto unpublished and all of types not reported by Godard 1950, or by
Wilkinson in his previous publications, were incorporated into the Ziwiye
repertory.
Wilkinson 1975 was not to be the last publication of Ziwiye ivories, for
K. Ishiguro (1976, Nos. 120, 147) published two from his collection. No. 120 is
a fragment of a winged sphinx in the round (of ink well type?); No. 147 is
an ivory chape. The former is a type hitherto not attributed to Ziwiye, while
the latter is of Achaemenid style, and also not hitherto attributed to Ziwiye.7
In the meantime Ghirshman had published his book on Iran for the Pen-
guin series (1954), wherein he discussed as deriving from Ziwiye some of
the objects previously published by him in 1950, still referring to the col-
lection as The Treasure of Sakkez, and not mentioning the name Ziwiye.
He did not mention a few of the objects he published in 1950 (figs. 4, 18,
19, 22, 23), which had been challenged by Godard 1951, but he did include
the gold bracelet and belt that were published by him in 1950 and that also
had been challenged by Godard; and he said nothing about fibulae or more
important, about ivories. In passing he published a terracotta ram-headed
situla (112, pl. 13b), attributing it to Azerbaijan although it had been previ-
ously published by Godard (1950, 68, fig. 57), as des environs de Ziwiye.
Later (Ghirshman, 1961, 100, No. 610, 1962, 75, fig. 20), ignoring his previous
designation, he moved the vessel back to Ziwiye.8 Ghirshman also contin-
ued to maintain that Scythian writing existed on the silver plate, a position
first advanced in 1950, but dismissed by Godard 1951 (244245, a reference
never cited by Ghirshman).
Godard had little to say regarding Ziwiye after 1951 (see Godard 1962,
9596), but Ghirshman continued over the years to write on the subject
and to introduce new objects into the repertory. Many other scholars had
contributions to make regarding Ziwiye and objects said to derive from there
and they will be discussed shortly.
Inasmuch as Ghirshman has been involved in the publication of Ziwiye
material for 25 years, one may follow the basic development of the modern
history of Ziwiye in his works. In 1961 Ghirshman played the major role in
7 See R.D. Barnett, Catalogue of the Nimrud Ivories (London 1975) T9, for a parallel to the
Ishiguro ivory No. 120. I believe that the ivory chape should be tested before it is accepted as
ancient; the eyes and ears are not similar to those noted on excavated examples.
8 This situla is apparently now in the Teheran Museum; it was originally in the possession
of the dealer Rabenou: Godard 1948, No. 8. The other example published by Godard, 1950,
fig. 58, as in his own collection, a fact not mentioned: but see Amiet 1969, 335336, figs. 19,
20, where it is stated that it was given to the Louvre by Mme. Godard. Note also that Amiet
says that the vessel came from Kaplantu, not Ziwiye, as was claimed by Godard (see note 19).
Tuchelt 1962, 58, No. 6, thought that the Godard situla was in Teheran. Keeping a record of
the modern proveniences of published Ziwiye material is not one of the easiest tasks for
those interested in this problem.
organizing what was presented as an exhibition of ancient Iranian art (see

Melikian 1961); it first opened in Paris and later travelled to other European
cities. The Paris catalogue (7000 Ans dArt en Iran) contained 1167 entries
of which 130, consisting of about 200 objects, were represented as deriving
from Ziwiye (Nos. 490624, omitting Nos. 490493, 509, 524, said to derive
from Kaplantu or elsewhere in Iran). Ghirshman wrote the entries and notes
for all the alleged pre-Islamic material including the Ziwiye section, called
Le Trsor de Ziwiy, Sakkiz having quietly disappeared.
Some of the material published in the Paris catalogue was familiar from
previous publications, and in a few instances additional fragments or units
of these pieces were made known. However, Ghirshman introduced for the
first time approximately 60 new types of objects, or varieties of types of
objects hitherto published, that he claimed without documentation derived
from Ziwiye. These were as follows: fragments of two gold epaulettes (500A,
pl. XXXVII; see Terrace 1966, No. 53); gold and silver bracelets (503, 504,
539); a gold chain (507; possibly Ghirshman 1964a, fig. 151); a gold glove (508,
pl. XLIV); a gold plaque (527); gold vessels (528531); gold elements (532);
gold torques (534536); gold necklaces (537, 538), a gold fibula decorated
with lion heads (543); a gold capital (546); a gold belt terminal (547; same
as Ghirshman 1964a, fig. 144); gold bracteates (548; see Ghirshman 1964a,
figs. 532537); gold earrings and rings (549; see Ghirshman 1964a, figs. 153,
529, 530); a gold seal frame (551; same as Ghirshman 1964a, fig. 140); a silver
pipe (552; same as Ghirshman 1964a, fig. 169); an iron dagger with lion head
protomes (555); a stone mace (556, pl. XLVIII); fragments of a gold cuirass
(557, 558); a silver shield fragment (559; same as Ghirshman 1964a, fig. 163);
a bone pommel (560); a bronze horse head (562, pl. XLVI); a bronze attach-
ment (568; same as Amandry 1966, 125, pl. XXVIII, 1ad); a bronze bell
(582); a bronze psalion (583);9 a bone harness attachment (584; same as
Ghirshman 1964a, fig. 540); 10 ivory and bone plaques and figures in addition
to those already published (587, 590, 591, 595, 596, 597, 599602; one same as
Ghirshman 1964a, fig. 136); a wood statuette (624; same as Ghirshman 1964a,
fig. 173); a gold plaque with a procession of men (595); a bronze lamp (603;
see Ghirshman 1964a, fig. 539); a stone lions foot (604); a bronze furniture
9 One of the many frustrations encountered by researchers of the Ziwiye problem is
inconsistency, as should be clear from a reading of the text. Thus Ghirshman 1964a, 120, says
only one element of psalia has survived, i.e. his fig. 540, and 1961 no. 584. He therefore
ignores his own 1961. No. 583 reference and also Godard 1950, fig. 46 (which he does not
otherwise challenge as deriving from Ziwiye); and note that Ghirshman 1973 cites an iron
example in Teheran; see also Porada 1965, fig. 73: and Jettmar 1967, 223!
piece (605); a bronze sphinx (606; same as Amandry 1966, 126, pl. XXVIII,
2ac); a terra cotta gazelle vessel (608; apparently not the same as Samedi
1960, fig. 34); a terracotta boot goblet (611); other terracotta vessels (612617,
620, 621); and fragments of terracotta sculptures (618probably same as
Ghirshman 1964a, fig. 172, 619). Note that some of these objects are in the
Teheran Museum, while others were in the hands of antiquity dealers (Coll.
particulire) at the time of the publication.
The pattern established in 1950 of publishing material claimed to be from
Ziwiye did not change, rather it was reenforced: no documentation or expla-
nation to underpin the introduction of new objects was deemed necessary;
one was apparently expected to accept the attributions automatically as an
expression of faith in the implied secret knowledge of the publisher. And
that the knowledge of both the objects existence and its alleged source came
from a dealer was not revealed.
The exhibition travelled to Essen in 1962 and a new catalogue was issued
(7000 Jahre Kunst in Iran). The Ziwiye section was reduced in size as some
objects were removed by their owners and no additions to the Ziwiye reper-
tory were made. The exhibition next moved to Zurich where a new catalogue
was issued (Kunstschtze aus Iran); many objects not published in the Paris
and Essen catalogues were added in a Nachtrag. In the Ziwiye section of this
Nachtrag 14 objects never hitherto published were introduced; several were
in the hands of dealers. They were as follows: a stone bulls head (847); a gold
protome (848, fig. opposite page 40); two gold beakers (849, 851, pls. 53, 32);
a fragment of a gold beaker (850); a gold cow (852, pl. 14); two gold goats
(853, pl. 15); a gold stag (854); a gold chain (855); two gold plaques (856, 857);
an ivory plaque (858); an ivory head (859); and a gilt bronze diadem (860).
Only five of the objects were illustrated, and none of these, I believe, should
be considered to be the work of an ancient artist (see below); inasmuch as
the others are not illustrated no comments can be made.
Of special concern with regard to the Ziwiye problem is that following
upon the massive intrusion of new Ziwiye entries in the Paris catalogue,
within a few months 14 additional objects surfaced from dealers and col-
lectors and were rushed into print and placed on view in the Zurich exhibi-
tion.10
In 1964 Ghirshman published The Arts of Ancient Iran which in-
cluded a chapter on The Scythians and the Royal Tombs of Ziwiyeh. Here
10 On page 9 of the Zurich catalogue Ghirshman, K. Erdman and P. Amandry are listed as
responsible for the additions.

Ghirshman presented his personal views concerning the nature of the trea-
sure, of which more later. He also introduced more objects into the Ziwiye
repertory, again without documentation: an ivory sword hilt (fig. 155; on
p. 425 it is claimed that a gold pommel with ivory lion heads belongs to this
piece); a second example of a sword hilt (fig. 161); a gold pommel (fig. 158;
cf. Ghirshman 1950, fig. 7; 1964, fig. 157); gold and silver weapons (100); seven
iron spear heads (100, 119, fig. 166); terracotta shields and lion-headed omboi
(100, 321, fig. 392), 11 bronze bells (120; cf. Ghirshman 1961, no. 582, where only
one is mentioned); stone vases (123); silver bracelets (113); terracotta animal
headed vessels (figs. 395, 396); and bronze and bone trilobate arrows (119,
fig. 165; see Dyson 1963, 37).
A number of photographs published in this 1964 work were of objects first
cited by Ghirshman in 1961 and earlier, but inexplicably no mention of this
fact occurs and one must painstakingly check each object against previously
published material.11 Moreover, both the descriptions and measurements
of what one believes to be the same object in separate publications vary,
which makes it difficult to be sure whether one has recognized correctly
the same object in its multiple publications, or there are two separate but
similar objects under discussion. It should also be noted that relatively
few of the countless objects in the Ziwiye repertory have been published
with a photograph. Therefore, it is possible that in some instances my
claim that an object is being cited of published for the first time may be
incorrect.
The problem regarding the ability to recognize whether or not an object
had been previously published, and the problem regarding the introduc-
tion and publication without documentation of material claimed for Ziwiye,
is succintly exhibited in Ghirshman 1973. There 341 objects in the Teheran
Museum, and claimed to be from Ziwiye, are published without illustra-
tions and in simple list form. No information concerning the dates and
sources of the acquisitions of the objects is given.12 Of the 341 objects I
11 In the text in my discussions of the 1961 material I add the 1964a references where I
recognize them.
12 Many scholars assume that if objects said to come from a given site are in the collection
of a Near Eastern national museum that the objects were scientifically excavated. But officials
of these museums often purchase objects from peasants or dealers, and, like their colleagues
in the West, supply provenience labels supplied by the vendor (Muscarella, 1977). Samedi,
1960, 18, for example, specifically states that the Teheran Museum purchased the Ziwiye
material; so also state Dussaud 1951, 287 f. and Bussagli 1956, 119; see also Melikian 1961, 67.
How many objects published by Ghirshman 1973 were acquired in 1950 and how many were
acquired in subsequent years is at present unknown outside of Teheran.
have been able to conclude tentatively that over 121 entries (some of which
represent more than one specimen) have apparently not been previously
published, that approximately 32 may have been previously published, and
that apparently the others have been published before.13 Given the meagre
published descriptions these figures may not be accurate. In 16 instances a
footnote indicates a reference to a previous publication but, strangely, in
other instances, although it can be determined that the object had been
previously mentioned in print, no citation or reference is presented. And
in the 32 cases where it is not clear that the objects have been previously
published, it is the apparent inconsistencies in comparison with the descrip-
tions and measurements of previously published pieces that cause one to be
confused.
Immediately after Godard 1950 and Ghirshman 1950 were published, the
scholarly world naturally became excited by the reports and over the years
often referred to the material and incorporated it into their studies. In
addition to the ongoing publications of Ghirshman and those of Wilkinson
cited above, more than a score of scholars had something to say about
the significance of the Ziwiye treasure in the history of 1st millennium bc
Near Eastern art, the various artistic strains recognizable, the chronology
of the deposit and of individual objects, the nature of the deposit, and
so forth. Some of these writings over the next decades were concerned in
varying degrees with some or all of these specific problems (viz. Falkner
1952; Frankfort 1955; Barnett 1956, 1962; Amandry 1958, 1964, 1965; vanden
Berghe 1954, 1959; Parrot 1961; von Ufford 1962, 1963; Porada 1965; Farkas
1970; Mallowan and Herrmann 1974), while others incorporated the alleged
Ziwiye material into their work on other aspects of ancient art (viz. Grousset
1951; Wiedner 1951; Benson 1960; Tuchelt 1962; Muscarella 1965; Culican 1965;
van Loon 1966; Jettmar 1967; Azarpay 1968; Moorey 1971; Hrouda 1971; Burney
13 I will spare the reader a listing of each of these objects except to give their numbers.
Those I suggest were not previously published include at least Nos. 10, 14, 18 (apparently not
the same as Ghirshman 1961, Nos. 534, 535), 1922, 30, 31, 3638, 43, 110 (cf. Godard 1948, No. 7;
Ghirshman 1964a, 113), 131, 132, 152159, 170, 202204, 252286, 288292, 300303, 305307,
312, 313, 320322, 325336, 355366, 410, 501, 503, 505, 506, 513, 514, 531 (apparently not the
same as Ghirshman 1964a, fig. 395, which is not mentioned in the present listing), 538, 539,
603608, and possibly also 507512, 515529. Those that may have been previously published
include Nos. 20 (cf. Ghirshman 1961, No. 504), 23, 25 (could be Dimand 1950, 145, left), 32 (cf.
Ghirshman 1961, No. 534), 161168 (cf. Godard 1950, figs. 167, 168, 169, 207219 (some of these
appear to have been published by Godard 1950 and Ghirshman 1964a), 314, 315, 348351; 532
may be the same as Samedi 1960, fig. 32.
1972; Phillips 1972; Calmeyer 1973). And still others continued the accepted
pattern of introducing new material by publishing an object acquired from
a dealer as automatically deriving from Ziwiye.
For example, Kantor published, in addition to the Cincinnati Museums
fragment of a trapezoidal plaque (1957, fig. 3), four other objects hitherto
not known to scholars: two fragments of gold plaques both depicting heros
killing lions (figs. 1, 2; see also Goldman 1964, figs. 9, 10); a gold and stone
bead necklace (fig. 4); and a gold bracelet (fig. 5). While I can say nothing
with regard to the authenticity of the last two items, on the basis of stylistic
analysis I believe that there is little reason to accept the first two as ancient
(cf. infra). What is more, in 1960 Kantor published a hitherto unknown
rectangular gold plaque depicting winged creatures (pl. I) acquired by the
Oriental Institute of the University of Chicago. To my eyes the details of the
creatures deviate considerably from those on the Ziwiye plaques to which
the Chicago plaque is compared; it equally deviates in style from other,
genuine, well-established, works of ancient art (cf. infra).
In 1957 Dyson published the University Museums alleged Ziwiye mate-
rial: more fragments of the ibex, stag, en face lion head plaque (fig. 25); a
fragment that is part of Godard 1950, fig. 29 (fig. 26). Also, for the first time,
he published a gold belt (?) fragment (fig. 26 top; cf. Ghirshman 1961, no. 518,
pl. XLI, 2, there said to be the same as Dysons fig. 26, but it is clearly another
fragment); a fragment of a gold plaque (fig. 26 right; cf. Ghirshman 1961,
no. 512, pl. XL, 3, from the Kofler-Truniger collection); and a gold bracelet,
formerly in the Kevorkian collection (fig. 28). Another fragment of a gold
plaque in the possession of the University Museum, and said to be from
Ziwiye, was published on the cover of the Guide to the Collections The Univer-
sity Museum (1965); on the basis of style, I am not convinced that it is ancient
(cf. infra).
Vanden Berghe 1959 (115, pl. 144 a) cited a bronze votive axe that he
claimed had been recently discovered at Ziwiye, but did not inform his
readers who made the discovery.
In Porada 1964 a gold funnel was introduced as being part of the treasure
(no. 437; it is said to be in the Teheran Museum but was not mentioned by
Ghirshman 1973; v. Calmeyer 1973, 143144). And Porada 1965 introduced as
from Ziwiye an ivory statuette of a male now in the Cincinnati Art Museum
(125127, pl. 35), and a bone psalion (fig. 73).
In 1966 the Muse Rath in Geneva held an exhibition of what it called
Trsors de LAncien Iran. Among a number of objects on view listed as from
Ziwiye, at least eight appear not to have been previously published: a gold
whetstone (No. 598, pl. 50); a gold horned griffin (No. 599); gold lion masks
(No. 600); a gilt bronze band (No. 603); a bronze mirror (No. 604); ivory
pyxis fragments (No. 605); an ivory plaque (No. 606); and an ivory human
mask (No. 607). Only one of these objects is illustrated and all seem to have
belonged to dealers.14
Finally, Amandry 1966 (121122, pl. XXV a, b, c) published three gold
fragments of plaques decorated with winged lions and human figures, which
he had seen on the art market. To my mind, all three appear not to be
products of ancient artists, a possibility hinted at by Amandry (see infra and
note 21).
2. Ziwiye/Zibia
Because of the phonemic similarity, Godard (1949, 168; 1950, 57) professed
that he had recognized in the name of the village of Ziwiye its ancient
name: que lon peut identifier avec lancienne Zibi, mentioned by Sar-
gon II (722705 bc) as a Mannean fortress. Many scholars have either with-
out reservation or tentatively accepted this equation: Dunand 1957, 287;
Wilkinson 1955, 213; 1967, 5 (but doubts raised in 1975, 9); Barnett 1956,
111112; Needler 1957, 9; von Ufford 1962, 38; 1963, 101; Boehmer 1964, 19
20, 21; Culican 1965, 42; Hrouda 1970, 249; Burney and Lang 1972, 135; Mal-
lowan and Herrmann 1974, 55. Aside from Wilkinson 1975, who pointed out
that the name Ziwiye is an Arabic word that could have been given as a
place name long after the fortress had been destroyed (sic) and for-
gotten, only Levine (1974, note 108) expressed reservations about the equa-
tion.
Ghirshman, on the other hand, laid stress on what he considered to be
significant, namely that Ziwiye was near the town of Sakkiz (it is actu-
ally about 40km. to the east of that site), and he emphasized what was
important about this proximity. To Ghirshman (1950, 201206; 1954, 106107;
1964a, 98, here with some reservations expressed), Sakkiz is to be iden-
tified with the Scythian capital of Partatua and his son Madyes. (It will
be recalled that in his earlier writings Ghirshman referred to the finds
as le Trsor de Sakkez, only later citing them as from Ziwiye.) Frank-
fort (1955, 205) accepted Ghirshmans identification, while Schefold (1954,
14 I can give no opinion here regarding the authenticity of these objects, as I saw them too
long ago. I am not sure if No. 598 is genuine or not (Muscarella 1977, note 81).
428) queried it Sakkez (stecken im Namen die Skythischen Saken?)

, and Dimand (1950, 145) used Saqqiz and Zeiwya (sic) interchange-
ably.
3. The Nature of the Treasure
Whereas Godard (1950, 910) had no hypothesis regarding the specific

nature of the alleged deposition beyond a suggestion that it was a treasure
cached by a king, a lord, or a plunderer, Ghirshman produced an elabo-
rate reconstruction of the events that led to the deposition. He did not
present his reconstruction as an hypothesis, but rather as an archaeological-
historical fact. In 1950 (202206) he connected the treasure to the relation-
ship between the Assyrians and Scythians, and its deposition to la dbcle
scythe de 625. He was more cautious in 1954 (106) when he stated that it was
probably a royal treasure which, we conjecture, may once have belonged to
Partatua or his son Madyes, but it is not known whether it came from a lomb
or a cache (italics mine). However, by 1960 (551) and 1961 (81) he apparently
came to know something, for he could now state that the find was Scythian
and derived du tombeau dun de leurs princes, and, further, that the trea-
sure accompagnait le prince, enseveli avec son entourage et ses chevaux
sacrifis.
In 1964a (99, also see pp. 111, 113, 125, 321) Ghirshman further expanded
his conclusions: there seems little doubt that the Ziwiye Treasure came
not from a cache, a crevice in the rocks where some robber lord concealed
the hoard, but from the tomb of a great Scythian king, buried in strict accor-
dance with the Scythian tradition; an appropriate quote from Herodotus on
Scythian burial customs followed. It seems that the documentation for these
conclusions was derived from an archaeological analysis of the types of
objects Ghirshman believed were found in the treasure (100): The remains
of a bronze sarcophagus prove that the Ziwiyeh hoard came from a tomb;
the gold and silver objects were probably the kings personal belongings.
But some ornaments are specifically feminine; The weapons in gold and
silver must have belonged to the king, but the seven long iron spear heads
suggest that the kings guards were immolated and buried with him; the
pottery must have formed part of the grave goods of sacrificed servants [see
also p. 123]. The possibility that a chariot and team of horses were buried
with the king cannot be excluded, in view of the presence of eleven small
bronze bells, miscellaneous horse-gear, and bronze plaques which may have
been affixed to a vehicle [p. 120 also]. He also argued (101) that the vertical
plaques of gold and ivory [were] used to decorate an article of furniture,

perhaps a throne or a ceremonial bed. Ghirshman repeated most of his
comments again (1964b, 6), but he further spelled out what the presence of
articles of jewelery in the tomb signified: they permettent d admettre que
les femmes ou les concubines sacrifies devaient suivre leur mari dans sa
derniere demeure.
Although Ghirshman since 1960 considered the treasure to be related to
a Scythian king, he also referred to the Scythian prince of Ziwiyeh (1964a,
104, 111), and still later in the same text (123) to the Scytho-Median princes;
this latter attribution was used again in 1964b (6) and 1974 (37): tombeau
princier mdo-scyth de Ziwiye. Thus, in reviewing Ghirshmans writings
over the years we note that the nature of the Ziwiye finds shifted from a royal
treasure related to the Medes in some manner (1954) to a burial related first
to the Scythians, and then to Median-Scythians (1960, 1961, 1964a, b, 1974);
and, moreover, that the burial was deposited replete with chariots, horses,
guards, servants, and wives or concubines of the king/prince.
The conclusion that the Ziwiye treasure derived from a tomb was as-
sumed also by Samedi (1960, 21); Barnett (1962, 94); Huot (1965, 139); van
Loon (1966, 177) and Orthmann (1975, 93). Barnett, with some modification,
suggests that the deposition represents not a Scythian but a Median chief-
tains burial ca. 600B.C. Orthmann avoids ethnic labels, claiming merely
that it was the tomb of an einheimischen Frsten. In addition, van Ufford
(1963, 101) not only accepted Ghirshmans conclusions totally, but embel-
lished them, making it clear to scholars that the Scythian kings wife (note
that this was written before Ghirshman 1964a and b) was also buried at
Ziwiye. Moreover, van Ufford ingeniously carried Ghirshmans conclusions
a step further in her attempt to elucidate the nature of the deposition.
She claimed that although the treasure indeed represented material from
a Scythian burial, what was in fact found were the remains of the burial that
had been previously plundered, and that the brigands ont rassembl leur
butin dans la cuve en bronze; puis ils ont cach (101). This archaeological
information was deduced because to van Ufford no Scythian would allow
himself to be buried in an Assyrian sarcophagus: of which, she said in the
previous paragraph, il nest pas mme certain que la cuve provienne du
tombeau.
Other scholars saw the deposition as a purposeful hoard or cache, with
no indications that it represented a burial (Porada 1965, 123124; Jettmar
1967, 222; Moorey 1971, 260; Phillips 1972, 135; Calmeyer 1973, 96, note 67).
A few scholars cautiously maintained that one simply did not know what
the deposition represented (Busagli 1956, 117, 119; Melikian 1961, 72; Dyson
1965, 207; Muscarella 1966, 381). Amandry (1965a, 897; see also 1965b, 150),
although accepting the possibility that le trsor accompagnait un mort
dans la tombe, pointed out (contra Ghirshman) on ne connait ni des con-
ditions ni mme Ie lieu de la trouvaille. La restitution de rites funraires,
fonde sur les objets retrouvs (sacrifice de chevaux, d esclaves ou de con-
cubines), comporte une grande part dhypothse. Au demeurant, les Scythes
n taient pas seuls immoler des chevaux dans les tombes. This provides
a neat summary of the situation, indeed, although the last sentence implies
that horse bones were found at Ziwiye, which is not the case, as Moorey
(1971, 260) pointed out.15
4. The Art of the Treasure
Given the vast number of objects attributed by many scholars to Ziwiye,

it was inevitable that the art historians among them would attempt to
define and isolate the various artistic traditions and elements reflected in
the material, and to draw cultural conclusions from their analyses. Godard
(1950, 11, 4452, 6677) saw the group, the Assyrian-style ivories excepted,
not as an ensemble gathered from different artistic and cultural centers,
accidentellement runis, mais reprsente lart pratiqu en pays mannen,
with, of course, foreign influences; the art of Ziwiye represents the art of the
Mannaeans. Godard distinguished three basic artistic elements represented
at Ziwiye: indigenous Mannaean, Assyrian, and Scythian.
Ghirshman at first (1950, 181; 1954, 106107) saw only four series of objects,
or distinct groups, Assyrian, Scythian, assyriennes lments scythes, and
a local group. In 1961 (8182) he added Luristan (following Falkner 1952 and
Amandry 1958), and Urartian (following Barnett 1956) elements; in 1964a
(100) Babylonian, Median, Cimmerian, and perhaps Greek elements were
further recognized; and in 1964b (7), he stressed the Urartian contribution.
Barnett (1956, 112) and von der Osten (1956, 57) were the first to call
attention to Urartian influences; later Barnett (1962, 91, 94) interpreted the
combination of Scythian and Urartian artistic elements at Ziwiye as signi-
fying that it was Median art, seeing the treasure, as mentioned above, to be
part of a Median chieftains burial. Falkner (1952, 129, 132) was the first to
15 Moorey also stated that no weapons were reported, which seems to be an error as
Godard, Ghirshman and vanden Berghe have cited them; and Dyson actually excavated three-
flanged arrows there.
mention Luristan art and Median and Syro-Phoenician elements, to be fol-

lowed by Ghirshman and later in part by von der Osten (1956, 57). Frankfort
(1955,205) and Sulimirski (1954, 298) simply accepted Ghirshrnans four cat-
egories. Kantor (1960, 68, 1013) emphasized Urartian influences; in fact,
seeing much of the treasure to be products of an Urartian tradiation; she
minimized, even eliminated, the designation Mannaean to describe the art
of the treasure. Huot (1965, 140) and Orthmann (1975, 93) also stressed the
importance of Urartian art as an influence on the objects. This latter concept
was vigorously seconded by van Ufford. However, succinctly demonstrat-
ing the problems with regard to who made what at Ziwiye, and how these
problems were treated by art historians, she first claimed (1962, 34) that
the gold gorget was made by a Luristan artist copying an Urartian model,
then later (1963, 105) claimed, without referring to her previous statement,
that it was made by a neo-Hittite artist who copied an Urartian model! The
concept that at Ziwiye there is evidence to indicate that artists from one
particular culture copied the style of, or fulfilled commissions for, another
was also maintained by van Loon. In 1966 (177) he asserted that Most of the
gold work is in a provincial Assyrian style, and that the hybridization
at Ziwiye suggested that one could recognize there components which he
tentatively calls Syro-Assyrian, i.e. perhaps made by a Syrian in Assyrian ser-
vice. But there was also another component at Ziwiye, its Syro-Urartian
counterpart, i.e. perhaps made by a Syrian in Urartian service. And there
were also pure Urartian and pure Scythian elements in the treasure. In 1967
(23) he emphasized the Syrian role in the art of Ziwiye, which suggested that
craftsmen from Syria having served various foreign courts, are responsible
for the [Ziwiye] pieces. Earlier, Frankfort (1955, 207) thought that local Ira-
nians copied Assyrian motifs for Scythian masters.
Schefold (1954, 439) preferred to stress only the Scythian elements, the
echt skythischen Funden von Sakkez. Porada (1964, 26; 1965, 130, 132, 238,
note 13), recognizing many artistic elements, nevertheless considered the
treasure to be of local production. She tentatively accepted Barnetts con-
clusion that the treasure represented Median art, but ultimately preferred
to reserve judgement on the lable to be given to the Ziwiye treasure until
more historical and archaeological data became available. And whereas in
1965 Porada noted Phoenician, North Syrian, Scythian, Urartian, and Ira-
nian elements to be present in the treasure, in 1975 (374) she emphasized
the Assyrian contributions. Finally, Culican (1965, 113, 118, 136138) accepted
Barnetts theories completely.
5. The Chronology
Because of what appeared to him as stylistic parallels with the art of 9th
century bc Assyria, Godard (1949, 169172; 1950, 10, 3644, 45, etc.) dated the
treasure to that time, singling out for special discussion the gold objects,
the gorget, trapezoidal plaque, the gold bracelet and some of the ivories;
other ivories he dated to the 8th century bc. The date of the deposition of
the treasure, as distinct from the date of the objects, he placed in the late
7th century bc (1950, 123).
Ghirshmans conclusions, published the same year as Godards book, and
citing some of the same objects singled out by Godard, were that on the basis
of style and because of the presence of fibulae, the treasure could not be pre-
8th century bc in date, and that it was probably to be placed in the second
quarter of the 7th century bc (Ghirshman 1950, 197198, 202); the deposition
was given as ca. 625bc, at the time when the Scythians were forced out
of Iran by the Medes. Ghirshman maintained these dates in subsequent
publications. Godard (1951, 242) reacted immediately to Ghirshmans low
dating for the treasure, rejecting the chronological conclusions based on the
fibulae (but, as noted above, neglecting to explain why he did not mention
their existence: see Muscarella 1965, 238, note 39).
In the many subsequent discussions of chronology, only two scholars to
my knowledge accepted Godards high dates for the manufacture of some
of the Ziwiye material, Parrot (1961, 142, 144), and von der Osten (1956, 57),
while a large number accepted Ghirshmans low dating: Sulimirski (1954,
313314), Wilkinson (1955, 218; 1960, 219220), Barnett (1956, 112; 1962, 94),
Needler (1957, 9), Amandry (1958, 16), van Ufford (1962, 38), van Loon (1966,
177). Other scholars opted for a general 8th7th century date (Kantor 1957,
14; 1960, 1), Huot (1965, 140), or a more specific late 8th century designation,
Hrouda (1970, 222, 249), Moorey (1971, 212, 260), Calmeyer (1973, 96, note 67),
and apparently also Mallowan and Herrmann (1974, 56). At least five schol-
ars preferred to recognize one or more objects as having been made in the
6th century, or preferred 7th6th century general dates, Falkner (1952, 131
132), von der Osten (1956, 57), Azarpay (1968, 46), Schefold (1954, 428), and
Benson (1960, 64, note 32); at least one of these scholars, Schefold, believed
that some of the material represented an early stage of Acnaemenid art.
Intimately linked with the problems involved in establishing an accurate

chronology for the Ziwiye objects is the question concerned with the origins
of Scythian art. In particular, the problem involves those objects that are
Scythian or exhibit Scythian motifs: the gold gorget, the gold plaque with
recumbent stags and ibexes and en face lion heads, the gold ibex, several gold
strips and buckles, the silver inlaid plate, the stone seal and the bronze and
bone horse psalia. And it is in this particular area, as much as any involved
in the Ziwiye problem, that the true provenience and chronology of the
alleged Ziwiye material is of great significance in the history of ancient art.
For if the objects mentioned above in fact derived from one site in Western
Iran, and if they were deposited together before ca. 625600bc, then they
legitimately couldshouldbe brought into a discussion involved with
the origins of Scythian art, which many authorities maintain manifested
itself in the Scythian homeland only during the 6th century bc, i.e. after the
alleged date of the deposition of the Ziwiye Treasure.
The theory that Scythian motifs and art were unformed and non-existent
until the Scythians came into direct contact with Iranian art during their
sojourn in Iran was first introduced by Godard (1949, 169170; 1950, 24, 44
49, 5964). Given the high 9th century dating of the Scythian material, and
given the much later appearance of Scythian art in Russia, Godards conclu-
sion was inevitable. In the same year that Godards statements were enun-
ciated, Ghirshman (1950, 202) came to the opposite conclusion. Because he
attributed the material to a lower date, one closer to the traditional dates
for Scythian art, he concluded that the Scythians actually brought their art
with them from their homeland and that the Ziwiye treasure was the earliest
extant examples of that art. Later, however (1961, 82; 1964a, 327), he reversed
himself (without calling attention to his previous opinion) and accepted
Godards theory (without citing him); his low chronology still gave priority
to the Scythian elements at Ziwiye.
Whereas certain scholars involved with the Ziwiye material avoided com-
ing to any conclusion regarding the origins of Scythian art, others expressed
their opinions, some vigorously. Sulimirski (1954, 313316), Potratz (1963,
97107, 111, 116), Amandry (1965a, 895896, 903904; 1965b, 150152, 159
160), and Phillips (1972, 136), supported the Godard theory; Burney and Lang
(1972, 175) simply stated that the earliest Scythian art is to be found at Ziwiye,
but drew no further conclusions. One of the first to oppose the Godard the-
ory was Falkner (1952, 132) who stated that Ziwiye alone in Iran produced
Scythian elements,16 and that it is strange that Mannaean elements appear
at the same time as the Scythian incursion. Barnett (1955, 114) also reacted
16 Scythian style bone psalia have since been excavated at Hasanlu from Period III: R.H.
Dyson, Jr., ILN, Sept. 12, 1964, 372, fig. 3 (incorrectly labelled as from period IV: see Muscarella
1974, 79, note 16); see also note 18.
negatively as he saw no resemblance between the local art of Mannaea and

Scythian art. Farkas (1970, 2024) went still further and challenged those
Soviet scholars who accepted Godards theory.17 She maintained that there
was in fact a long tradition of animal art in the Scythian and neighbor-
ing lands for centuries before the Scythians invaded the Near East. To her
the earliest known group of objects in Scythian animal style, the so-called
Ziwiye treasure must have been made by Near Eastern craftsmen who
combined Scythian elements with their own artistic traditions. In short, the
Scythian elements at Ziwiye were of Scythian, not Near Eastern origin; the
Scythians did not borrow artistic motifs from Near Eastern artists, they lent
them.
Van Loon (1970, 6970), echoing in part Amandry (1965a, b), was the
only scholar to my knowledge who rejected the Godard (and Potratz) the-
ory on firm archaeological grounds: This evidence [the Ziwiye material]
cannot carry much weight, as the context of the clandestine Ziwiye finds
is unknown and their various dates and countries of origin are debated.
The arguments would be stronger if excavated and definitely earlier proto-
types could be adduced for some class of Scythian objects.18 Wiedner (1951,
152) also expressed caution with regard to the arguments between Iranian
or homeland priority by stating that if Godards dating is correct, Scythian
art did originate in Iran, but he preferred to see more objects published
before drawing any concrete conclusions. Grousset (1951, 109110), vanden
Berghe (1959, 114), and Jettmar (1967,223: the matter is not so simple), also
expressed caution, refusing to come down for or against Godards theory.
6. Doubts
At this point in our itinerary through the Ziwiye maze the reader may feel
overwhelmed both by the investigation of various scholars opinions, embel-
lishments, conclusions, omissions, disagreements, and so forth, and by the
listing of the many hundreds of objects that have surfaced over the years
and that have been assigned to Ziwiye. It is now pertinent to ask if anyone
17 Viz. Artamanov, which reference I know from Farkas 1970 and van Loon 1970, 69, note 23.
(I belatedly came across S. Sorokin, The Curled-up Animal from Ziwiye, [in Russian] Soob-
schenia Gosudarsvennovo Hermitazha 34 [1972] 7578, where the gold pommel published by
Ghirshman 1964a, fig. 158, is used in an argument that the animal style of the Scythians had
its origins in the Near East.)
18 Unfortunately van Loon assumed that bone psalia of Scythian style had been excavated
from Hasanlu period IV: see note 16.

raised doubts or asked questions regarding the discreteness of the corpus,

about the sources of the information that allowed objects to be assigned
to Ziwiye, or about the validity of the alleged provenience in general. In
a few instances, when a specific issue was being examined in the previ-
ous discussions, questions raised by certain scholars have been mentioned
(viz. Amandry, van Loon). And an investigation of the literature reveals that
actually quite a few scholars did express doubts or reservations about indi-
vidual objects being included in the corpus, or expressed concern that deal-
ers could have salted the finds or simply misattributed material known to
have been derived in fact from other sites. But in practically every instance
where reservations or doubts were stated, the same individuals either them-
selves published new material as deriving from the site, or accepted objects
published by others, or both. Thus, in the former instance, it was assumed
that it was the other person who might be polluting the Ziwiye corpus, not
the writer himself, and in the latter instance, it was implied that the writer
knew, for reasons never vouchsafed, what objects did and did not derive
from Ziwiye.
Paradoxically, because its significance was never fully recognized by the
writer himself nor by his many successors, Godard warned his readers at
the very beginning of the Ziwiye odyssey about the problem of false prove-
niences, albeit burying the warning in a footnote (Godard 1950, note 128).
After first reassuring his readers that les pices que j ai prsentes ici
comme appartenant au trsor de Ziwiye lui appartiennent bien, he went
on to state cependant il est certain quil arrivera, comme il arrive toujours
en pareille circonstance, quapparatront bientt dans le commerce, comme
provenant de Ziwiy et du trsor mme, des objets d une tout autre origine
et dun tout autre temps. Inasmuch as Godard got to know all the objects he
published dans le commerce, it surely is not gratuitous to note that the first
sentence of the above footnote is not only misleading, it is meaningless. In
his arguments that some of the material published by Ghirshman 1950 did
not come from Ziwiye, Godard (1951, 240) repeated his warning, but only in
the context of his disagreements with Ghirshman. He did not consider that
his warning applied to himself nor to the bronze plaque, the silver dish, and
the fibulae published for the first time by Ghirshrnan, objects which the lat-
ter got to know dans le commerce.
In 1951 Dussaud reviewed both Godard 1950 and Ghirshman 1950. While
explicitly accepting Godards material as deriving from Ziwiye, he was per-
plexed by Ghirshmans material. He had been led to believe that Godard
had published the entire collection, which had been acquired by the
Teheran Museum, and which to him had a homogenous character, while the
Ghirshman ensemble had a clearly disparate nature (Dussaud 1951, 290). He

also posed a provocative question and issued his own warning: Une ques-
tion prjudicielle se pose: ces objets proviennent-ils vraiment du fameux
trsor? Lors de toute importante trouvaille fortuite, marchands et collec-
tionneurs sont enclins y joindre certaines pices plus ou moins similaires.
Les problmes que soulve le trsor, dment certifi par une enqute svre,
sont assez difficiles resoudre ne viennent pas y apporter un trouble irrm-
diable (Dussaud 1951, 290). It was a prophetic statement, among other
things, but written specifically contra Ghirshman 1950, and not intended to
apply to Godard 1950.
Frankfort (1955, 263, note 11), Amandry (1958, 17, note 59, echoing Dus-
saud: On sait que ces pices de diverses provenances ont t mles ind-
ment aux objets du trsor.), and Wilkinson (1963, 274275; 1975, 78) also
accepted without reservation the attributions of Godard 1950 but were on
guard against Ghirshman 1950; and the latter two scholars were not loath to
introduce and discuss other objects as derived from Ziwiye. We have seen
how van Loon (1970, 6970) and Amandry (1965a, 897) came as close as any-
one to expressing concern about using Ziwiye material for firm historical
and archaeological conclusions. Amandry (1965b, 150) further noted that
aucun observation ait t faite sur les conditions de trouvaille, ni aucun
inventaire tabli, but he himself accepted a great amount of the published
objects as deriving from Ziwiye, and in 1966 he also introduced new mate-
rial. Perhaps the most unexpected warning of all, given the vast number
of objects that he had introduced, is that of Ghirshman (1964a, 321) who
informs us that attributions of objects found in unscientifically conducted
excavations can never be wholly relied on. Nevertheless And he was writ-
ing about Ziwiye!
A few scholars who expressed reservations about certain objects attri-
buted to Ziwiye may be singled out. Dyson (1963, 34), although accepting
as fact that something was found at Ziwiye because a dealer worked there
for years after 1947, added whether all of the objects since attributed to the
find came from here or not is open to question; and in 1965 (207) he noted
that nothing is known of the circumstances of their discovery beyond the
assertion that they were associated with a large coffin. Porada (1965, 125) felt
obliged to note in her chapter on Ziwiye that one must always stress that
objects said to come from Ziwiye cannot be proved actually to have been
found there; and (1975, 374) that nicht einmal die Angabe der Herkunft-
sortes als unbedingt gesichert gelten kann. Mallowan and Herrmann (1974,
55) remarked that the collection has meretriciously attracted much else
that does not belong, and in some cases may not be genuine. They also
challenged the assumption that all the ivories attributed to Ziwiye derived
from that site: these ivories came not from one, but from many findplaces
(57); but at the same time they accepted a number of examples, especially
those published by Godard and most of those published by Wilkinson, as
in fact coming from Ziwiye. Moreover, as mentioned above, they accepted
with some assurance a Ziwiye provenience for objects acquired by the
Metropolitan Museum. And this writer has also occasionally called atten-
tion to the problems of attribution or provenance (Muscarella 1965, 233,
note 6; 1966, 380381; 1976, 210).
It must be stressed again, however, that all of the scholars who had
reservations of various degrees nevertheless at one time or another either
accepted some or most of the published material as coming from Ziwiye, or
in some cases themselves introduced new material as coming from there.
Thus, their doubts were primarily ad hoc ones, reservations against individ-
ual objects, or they were general ones, left isolated and not developed as a
problem probably effecting the whole Ziwiye corpus.
7. Kaplantu/Qaplantu/Ghaflantu
Inasmuch as the site of Kaplantu is often mentioned together with Ziwiye,

and objects are published as deriving from there, a brief discussion about
its modern history is warranted. Kaplantu was first mentioned by Godard
(1950, 67) as the possible site of ancient Izirtu. It was not made clear
why this village, about 5km. SE of Ziwiye, was thus honored except that
Godard mentioned, almost parenthetically, des sondages excuts dans un
ancien cimitire de lendroit ont livr des objets du mme temps et de la
mme qualit que ceux qui proviennent de Zibi. Presumably, one was
meant to accept the archaeological equation that sites allegedly yielding the
same quality of material and apparently of the same period must be both
contemporary and equally important. But Godard neither mentioned nor
described, let alone illustrated, a single object allegedly from Kaplantu, and
he did not explain how he knew of the cemetery.19 To date the site remains
scientifically unexcavated.
19 With no objective information in hand it may be rash to speculate on Godards source
of information regarding alleged finds at Kaplantu. Nevertheless, I believe it is probable that

Godard got his information from the same dealer who was his prime source for the Ziwiye
material. It is also possible that Godards term environs de Ziwiy may have been his way
of dealing with material he thought came from Kaplantu: see for example note 8; also see
Wilkinson 1975, 7.
In 1960 Ghirshman published three objects that he claimed came from

Kaplantu, at that time still in dealers hands: fig. 3, a silver rams head vessel
(now in Cleveland, Shepherd 1966, 4749, fig. 13); fig. 4, a gold finial (now
in the Nelson Gallery of Art); fig. 8, two gold goat handles (now in the Freer
Gallery). The next year, in the Paris Exhibition catalogue, Ghirshman pub-
lished as from Kaplantu, in addition to the Cleveland silver vessel (No. 492),
a gold rams head vessel (No. 490), a gold whetstone (No. 493), and a gold
disc decorated with a hero holding two lions (No. 509). Typically, no expla-
nation for the attributions was presented. Nor was any offered in Ghirshman
1962a, where he again published the gold rams head vessel, and another
silver rams head vessel, from the Pomerance collection (75, figs. 18, 19; in
Ghirshman 1961, No. 491, this vessel was simply listed as from Iran).
Eventually a total of five silver rams head vessels were published, all
said to have come from Kaplantu: the two already mentioned in Cleveland
and in the Pomerance collection, one in the Abegg collection, one in the
Metropolitan Museum of Art, and one in Philadelphia (Wilkinson 1967, 10
21, figs. 511, pls. IVII). Culican (1965, 122), Shepherd (1966, 4748), Terrace
(1966, 48, no. 54, with a question mark), and Seidl (1968, 345, b, c; listing
only two vessels of this type) also referred to these vessels as being from
Kaplantu. Wilkinson (apud Hoffmann 1961, 21, note 11, pl. 10:5) attributed
a terracotta bent rhyton from Kaplantu, near Ziwiye, with no reference.
Wilkinson (1967, 22, figs. 12, 14, pls. VIIIXII) also added to the list of the
rhyta from Kaplantu, a bronze situla in the Abegg collection, and a glazed
terracotta vessel in the Metropolitan Museum of Art (see also Orthmann
1975, 93, pl. XXXVIII). He justified the attribution of the silver and bronze
Abegg vessels to Kaplantu because the former has some obvious similarities
with the other silver rhytons that are attributed to the Qaplantu and Ziwiye
region, and the latter has links with vessels from that region (Wilkinson
1967, 13, 26). In other words, some vessels stated by dealers to have come
from Kaplantu and Ziwiye look like other vessels said by dealers to have
come from Kaplantu and Ziwiye. A few years later, Wilkinson (1975, 6970,
figs. 39, 39a) published a gold band in the Abegg collection as reputed to
come from Ghaflantu; and he illustrated (fig. 39b) a similar example in the
Metropolitan Museum of Art.
Over the years other stray objects were placed in the Kaplantu scrap
basket. A terracotta situla first published by Godard (1950, fig. 58) as from
the environs de Ziwiy, and at that time in his private possession (see
note 8), was later claimed by Amiet (1969, 335, figs. 19, 20) to have derived
from Kaplantu. The 1966 Muse Rath catalogue presented three new objects
as deriving from Kaplantu: a gold vase (No. 619, pl. 46), a gold bracteate
(No. 621), and eight stone ram heads (No. 623); also published here was the
gold rhyton of Ghirshman 1961, and another gold disc with a hero hold-
ing lions (No. 622). A gold finial in the Los Angeles County Art Museum,
formerly in the Berg collection, is listed as Found at Ghaflantou, Iran (Mus-
carella 1977, no. 168). And finally, in 1968 the Reallexikon der Assyriologie
und vorderasiatischen Archologie published its entry on Ghaflantou (Seidl
1968, 344346) wherein it is stated that wenigstens 13 figrlich und orna-
mental verzierte Fundstcke bester Qualitt aus dem I. Jt. sollen dorther
stammen. As for bester Qualitt I shall speak below; the sollen appar-
ently is meant to supply a clue that there is a lack of any facts regarding
provenience. But this latter intimation is unfortunately not pursued, and
one has the distinct impression that the reader is being informed about
material that came from Kaplantu.
8. Suspicious Objects
A. Ziwiye
We have seen how since 1948 a continuous stream of objects of various
materials said to be from Ziwiye has flowed into the literature and into the
exhibition cases. If a dealer, whether in Teheran, Switzerland, London, Paris,
or New York, said an object came from Ziwiye, this usually was considered
sufficient archaeological evidence by some members of the scholarly com-
munity to warrant an attribution. Given this abdication of scholarly respon-
sibility, it could almost have been assumed (as by Dussaud 1951) that objects
from other sites or from other periods would be inadvertantly included
within the Ziwiye corpus. And surely in the present case it is hardly cyni-
cal to note that given the total corpus attributed to the site, the single cuve
de bronze, accepted by most scholars as originally containing the treasure,
could never have held all the many hundreds of objects in existence; at least
three or four such containers would have to be posited!
Some scholars, as already noted, did call attention to the possibility of a
false Ziwiye provenience for stray objects, but very few discussed in print the
possibility that a false provenience might also provide a pedigree for forg-
eries or suspicious objects, objects that should never have been presented
as deriving from any ancient site, Ziwiye or elsewhere. To my knowledge
only the writer (Muscarella 1977, nos. 155165) and Mallowan and Herr-
mann (1974, 55) specifically mentioned the possibility that forgeries may be
present in the corpus. Melikian (1961, 71) has presented the problem in gen-
eral, but his statement aptly applies to the Ziwiye material: Rien n est plus
anonyme que lobjet de fouilles clandestines, et rien ne convient mieux au

faussaire que cet anonymat. On travaille encore admirablement le bronze
cisel en Perse, et lon sait tourner de fort belles poteries.
There are among the published material said to come from Ziwiye at least
17 objects that I consider to be suspicious and that I suggest should not be
accepted as ancient until proven to be so. The reasons for designating these
objects suspicious are not simply that they do not conform to the other
alleged Ziwiye material, for if this alone were the case the objects could be
genuine but from another site. It is because in style and detail, on the basis of
internal analysis, these objects do not conform to what we know occurs on
ancient, excavated, art. If we accept the majority of the objects said to come
from Ziwiye as genuine because they do conform in style to what we expect
on ancient objects, then by the same standards of art historical analysis we
have the right to challenge those that do not conform to what is expected (cf.
Muscarella 1977). The challenge to the objects singled out is not presented
in categorical fashion. Yet given the significant facts that, first, they have not
been excavated and, second, that ancient parallels are not forthcoming, they
cannot be automatically accepted as genuine merely because they exist. The
question to be asked is not so much why are they suspicious, but rather, why
are they genuine. Moreover, to the anticipated indignant question: how can
one tell that an object is a forgery or suspicious merely from a photograph?
there is a simple answer: how can one tell that an object is genuine merely
from a photograph?
I list the suspicious objects in the order of their publication: 1) gold
plaque, Metropolitan Museum, Wilkinson 1955, 216; Muscarella 1977, No. 156;
2) gold plaque, Cincinnati, Kantor 1957, fig. 1; Muscarella 1977, No. 163; 3)
gold plaque, Cincinnati, Kantor 1957, fig. 2; Muscarella 1977, No. 164; 4) gold
plaque, Oriental Institute, Univ. of Chicago, Kantor 1960, pI. 1; Muscarella
1977, No. 162 bis; 5) gold plaque, British Museum, Barnett 1962, pl. 1; Mus-
carella 1977, No. 165; 6) a possibly genuine ancient bronze bucket in the
Metropolitan Museum (Wilkinson 1963, figs. 14, 15; Muscarella 1977, No. 162)
has been embellished in modern times with a design consisting of two heros
fighting lions separated by two trees and bordered by guilloches (estab-
lished by laboratory analysis); 7) gold plaque, University Museum, Philadel-
phia, Guide to the Collections of the University Museum, cover; Muscarella
1977, No. 155; 8) gold plaque, Ghirshman 1961, No. 527; Muscarella 1977,
No. 165; 9, 10, 11) three gold plaques, art market, Amandry 1966 pl. XXVac.20
20 After publishing these plaques Amandry (122) suprisingly and without elaboration adds
La scheresse un peu mcanique du style invite la reserve.

Muscarella 1977, No. 165; 12, 13, 14, 15, 16) a gold ram, two gold vessels, a
gold cow, gold goats, Ghirshman, et al., (Kunstschtze aus Iran), Nos. 848,
849, 851853; Muscarella 1977, Nos. 157161; 17) silver horse frontlet, British
Museum, Barnett 1973, 125, pl. LIII a (which may be the same piece men-
tioned by Amandry 1966, 125).21 In addition, there are two other objects
whose authenticity or lack of same is not clear to me: a gold decorated strip,
Ghirshman 1950, figs. 18, 19 (cf. Muscarella 1977, No. 54); and a gold bracelet,
Cincinnati, Kantor 1957, fig. 4).
B. Kaplantu
From the relatively small group of objects listed as having derived from
Kaplantu, a good number, to my mind, do not compel one to accept them as
ancient, and they should be considered to be suspicious until proven other-
wise. As with the Ziwiye material singled out above as suspicious, analysis
of style and artistic motifs has been the criterion for judgement, not merely
a lack of apparent artistic relationships with material alleged to be from
Ziwiye or its neighborhood; all the objects have been mentioned above: 1)
gold rhyton, art market, Ghirshman 1961, No. 490; Seidl 1968, a; Muscarella
1977, No. 166; 2) gold vase, Geneva, Muse Rath No. 619; Seidl 1968, 1; Mus-
carella 1977, No. 170; 3) two gold goat handles, Freer Gallery of Art, Ghirsh-
man 1960, fig. 8; Seidl 1968, k; Muscarella 1977, No. 105 (to my mind actually
meant to be Achaemenian); 4) silver ram-headed vessel, Abegg collection,
Wilkinson 1967, 1214; Muscarella 1977, No. 167; 5) gold whetstone, Nelson
Gallery of Art, Ghirshman 1960, fig. 4; Seidl 1968, e; Muscarella 1977, No. 169;
6) gold whetstone, Los Angeles, ex Berg collection (same as Ghirshman 1961,
493 and Seidl 1968, f?); Muscarella 1977, No. 168; I also believe that the gold
discs depicting a hero holding lions deserve more technical and visual exam-
ination before they are all accepted as ancient, Muscarella 1977, note 86; cf.
Seidl 1968, g-i.
21 This object was not mentioned in Muscarella 1977 because I only recently learned that it
was in the British Museum and had been published; previously, I knew it as an unpublished
object in a dealers possession, and therefore could not mention it. It is very clear that the
British Museum example differs in execution and in details from the published frontlet of
Godard 1950, fig. 109: compare the spacing and size of the punched dots; the trunk, flowers,
and design of the trees, as well as their respective qualities of execution; and compare the
obvious differences in the nature of the body decoration, genitals, tail ends, mouth, eyes, and
feet of the two lions. I believe that a laboratory analysis on the British Museum example is
warranted. Note also for the record that Melikian 1961 and Lorenz 1970 should be added to
the bibliography of Muscarella 1977.
9. The Excavation of Ziwiye
From the evidence available in print or from private information, it may be

firmly stated that there never were scientifically controlled excavations at
Ziwiye22 until 1964, 17 years after the time of the alleged discovery of the
site. Yet a review of the published literature reveals that certain scholars
have misled both themselves and their colleagues by asserting or imply-
ing that controlled archaeological excavations existed at Ziwiye after 1947.
This deception occurred primarily because the term excavate was broadly
employed to describe the activities of dealers searching for treasure, and
this led to the supposition that some archaeologist must have worked there.
Unfortunately, no one bothered to ask: what archaeologist? In addition,
it seems that a large number of scholars were (and still prefer to remain)
innocent about the role of dealers and their commercial digs in Iran, while
others who did know, for reasons best known to them, chose to keep this
information a secret.
Godard himselfwho knew the factswas the first to imply that excava-
tions (in the correct sense of the word) had occurred: Les fouilles que l on y
pratique actuellement met tent jour. On y trouve aussi. (Godard 1959,
7). Compounding the misleading implications of his original statement, he
later (1951, 241) cited le fouilleur lui-mme as the one who gave him infor-
mation about specific find-spots at Ziwiye. The fouilleur, however, was not
identified, and remained so until mentioned by Dyson (1963, 34). To those
not familiar with the fact that it was a dealer who was excavating at Ziwiye,
the implication was that some objects, at least, if not those from the treasure,
had been scientifically unearthed.
Falkner (1952, 129) discussing Godard 1950 mentioned eine kurze Unter-
suchung, that produced walls and so forth; and Kantor (1957, 10), van-
den Berghe (1959, 115), and Samedi (1960, 24), employing archaeological
terminology, referred to recent tests, or to sondages conducted at the
site. Needler (1957, 9), for some unexplained reason, made a firmer state-
ment claiming that there has since [1947] been some controlled excava-
tion at the site; and Wilkinson (1967, 10) put forth the claim that certain
animal-headed vessels came from the first excavations in that area, i.e.
Ziwiye. All these statements made their contributions to the confusion
concerning the crucial distinction existing between digging and scientific
22 And here. of course, we are concerned with Ziwiye, without the mandatory quotation
marks, Ziwiye, that hitherto was assumed.

excavating, and between information available about the objects derived

from the two separate processes. Further reenforcing this confusion were
the occasional remarks by scholars to the effect that a particular object
from the Ziwiye corpus was indeed found or excavated there: viz. a horse
bit was actually found (Frankfort 1955,207); recemment on a dcouvert
une hache votie (vanden Berghe 1959, 115); a gorget was found at Ziwiyeh
(Muscarella 1971, 264).
Statements about the true nature of affairs at Ziwiye did turn up in
print albeit rarely. Kantor (1960, 1) was the first one to mention the crucial
fact that there were licensed commercial excavations existing at Ziwiye.
Later, T.C. Young, Jr. (1965, 59) mentioned the extensive trenches opened
by commercially licensed diggers; and Lorenz (1970, 3940) specifically
employed the word Raubgrabung to describe the finds at Ziwiye. Dyson
in his first report on Ziwiye (1957, 33) stated that excavations had been
conducted since 1947, but later (1963, 34) he corrected the vagueness of this
statement and specifically referred to and named the dealer who had been
digging for treasure at Ziwiye for years.
The first controlled excavation of Ziwiye was conducted in 1964 by a team

from the Hasanlu Project under the direction of R.H. Dyson, Jr. However,
Dyson first visited the site in 1956, which resulted in a brief report the
following year (1957, 3337). This report was based on first hand observa-
tion of the site, the first one of such nature ever published by a scholar.
Dyson mentioned seeing traces of architecture, floors and mudbrick walls,
stone column bases, and a portico; he also picked up copper/bronze frag-
ments, rivets, threeflanged arrowheads, triangular glazed tiles, and pottery
sherds, all courtesy of the trenches dug by the dealer who had preceded him
there.
Dyson returned to Ziwiye in 1960 and again conducted a survey of the site
(Dyson 1963). He reported observing three occupational areas terraced one
above the other, the highest containing fragments of large jars. In addition
to reporting on the same type of material recorded in 1957, he found and
illustrated painted sherds, a fragment of a terracotta lamp, fragments of
stone bowls, a broken bronze bracelet, a bronze fragment of scale armor,
another three-flanged arrowhead, and a bronze fragment decorated with a
man on a horse (Dyson 1963, 3537). This last object has no stylistic parallels
with any of the published material said to come from Ziwiye. It should also
be recalled that T.C. Young, Jr. conducted a survey of the site, making a
collection of sherds which he published with full descriptions (Young 1965,
5861, figs. 35).
The results of the three-week excavation undertaken in 1964 were re-

ported in summary fashion in Dyson 1965 (205206). A sloping, paved stair-
way, of which 30m. were cleared, led up to a fortified castle, a structure
which consists of a complicated system of corridors and small rooms sur-
rounded by a wall some seven and a half meters thick. Of some importance
is the conclusion that there were three separate phases or stages of rebuild-
ing at Ziwiye with no signs of destruction by fire. Also of interest, although it
is not yet clear if it is significant, is the fact that while fine incised ware sherds
were recovered from the excavations in the structure, painted ware was
recovered only in the surface debris. Dyson dated the site between 750bc
and 600bc, a date generally accepted by Young, Boehmer and Muscarella
(Muscarella 1973, 7172, notes 20, 21; Boehmer 1964, 20, note 68).23
10. Conclusions
The Ziwiye problem had its beginning with the first publications by Y. and
A. Godard in 1950. The research involved in digging through and record-
ing each successive level deposited in the ensuing mound of bibliography
has been very similar to the activity carried out on an archaeological field
campaignbut a campaign conducted on a mound where each level is dis-
torted by the existence of many pits, disturbed and missing areas, and com-
plex, non-uniform structural additions, alterations, and repairs; a campaign
which yields not a controlled horizontal sequence of events, but rather often
a convoluted, spiral-form stratigraphy. And a section drawing of the multiple
scholarly views that refer to specific find spots and dates, chronology, artis-
tic elements, the nature of the deposition, and which objects may or may
not have been found, and so forth, would form a series of fragmented loops
and spirals, crudely interlocking and overlapping one another.
One of the aims of this study has been to isolate these spirals and to
attempt to unwind them, to make them horizontal, so to speak, so that
23 In the course of writing this paper I learned that Nosratollah Motamedi of the Iranian
Center for Archaeological Research conducted an excavation at Ziwiye in 1976. A brief report
on the finds said to be from two cemeteries is given in the Exposition des Dernires Dcou-
vertes Archologiques 19751976 of the Fifth Annual Symposium of Archaeological Research
in Iran, November 2- December 2. 1976, 2729. with references to the material in vitrines 18a
and b. Among the objects mentioned being excavated at Ziwiye are terracotta vessels, a terra-
cotta lamp, bronze pins, bronze earrings, a bronze tweezer, an ivory fragment (!), frit cylinder
seals, a frit scarab, a spurred arrowhead, and a fragment of a terracotta with an enamelled
motif.
others may be able to form their own conclusions, to draw their own sec-
tions, if they choose to reject those offered here. Another aim has been to
set forth as much of the working data that is presently available in order
to assist in attempts to answer several important questions. Was anything
found at Ziwiye in 1947? What was found? Was anything found there after
1947 by commercial diggers? What historical and archaeological value does
the Ziwiye material have? To those questions we now turn.
Was anything found at Ziwiye in 1947? The alleged discovery of the trea-
sure is not challenged by anyone as having occurred fortuitement. Inas-
much as the first reported objects surfaced via the antiquities market, they
certainly resulted from a clandestine dig (one of the few undisputed facts
available in the Ziwiye problem!). However that came about, and whenever
it occurred, this fact must be the starting point of any investigation and dis-
cussion. Only the individuals involved in the clandestine find know where
the site was and what were the circumstances of the discovery. Therefore,
it follows that the Godards statements regarding the story of the Ziwiye
find, the list of objects recovered from the bronze container, and those
objects found outside it, did not derive from personal observation but from
second- and third-hand sources: from dealers and possiblybut on this
issue one can only guessfrom the Ziwiye villagers themselves. As such,
all the Godards statements and claims are hearsay, and are not what is tra-
ditionally considered to be sound sources of information in archaeological
research, scientifically controlled and recorded observation by an archaeol-
ogist and his staff.24
It is equally certain that Ghirshman knew nothing objective or first-
hand concerning which objects were actually found at Ziwiye, nor even
the date when the find was said to have occurred; he too knew only what
was revealed by dealersobviously in some cases dealers other than those
who talked to Godard! In 1950 the scholarly community was confronted by
the publications of Godard and Ghirshman, both works written apparently
independently and unaware of each other, both implying that a corpus of
the Ziwiye material was being presented, and yet each publishing material
not mentioned by the other, and omitting material given great publicity by
the other. Godards 1951 retort did not resolve these obvious contradictions,
rather it created more; the spiralling had begun. As reported above, Godards
24 In this context one need only recall the confusion, the bitterness, and the different
opinions held by the archaeological community regarding whether or not the so-called
Dorak treasure exists, merely because it was reported by an archaeologist working alone who
claimed to have seen it.
warning that dealers gratuitously assign labels to stray objects they acquire
was thrust forth solely contra Ghirshman, to score points in a specific argu-
ment, claiming that some of the Ziwiye material did not come from
Ziwiye.25 And because of the narrow intent of this warning, Godard did not
recognize it for what it actually was, a warning applicable to all material
offered for sale by dealers, which embraces all the material allegedly from
Ziwiyeincluding that published by Godard himself. Ghirshmans warning
regarding the attribution of dealers was also ad hoc, against other scholars,
and was not meant to reflect on his own enormous list of Ziwiye material, all
of which he maintained provenant dun seul et mme trsor (Ghirshman
1950, 198).
If the earliest writers on Ziwiye established a pattern concerning the
treatment of the material attributed there, most scholars in the follow-
ing decades uncritically accepted it, indeed refined and institutionalized it.
They too assumed the right to publish any object purchased from a dealer
and assign a Ziwiye label. Generally, without any critical evaluation, they
too expected automatic acceptance of their assertion, masking the undocu-
mented provenience with phrases such as said to come from Ziwiye, or
more emphatically stating that it is practically certain, or there seems
no reason to deny, that the object came from Ziwiye. Museums also con-
tributed in large part to the dissimulation and confusion concerning prove-
nience by allowing exhibitions of Iranian Art to serve as a vehicle for the
exhibition and publishing of newly introduced material alleged to be from
Ziwiye. Moreover, much of this material was in the hands of dealers at the
time of the exhibition and the museums thereby institutionally stamped the
objects both as from Ziwiye and as genuine. That endorsement helped to sell
them, courtesy of the documentation acquired.26
Are we able to learn anything about the alleged find from the villagers
at Ziwiye itself? Dyson (1963, 34) stated that in 1956 the local peasants vol-
untarily pointed out the spot where the chest had allegedly been found,
and that he believed they had no reason to lie to him. Further, commenting
on the commercial digging he witnessed that same year, he thought that it
was doubtful that the dealer would have been spending money excavating
25 There appears to have been a personal battle in progress between the protagonists
being worked out in print, and with archaeological knowledge as the victim.
26 Many of the objects in dealers hands on exhibition were subsequently sold to museums
and private collections: and always the vendor cited as evidence of authenticity and prove-
nience that it was exhibited in a major museum. These exhibiting museums thus functioned
both as bazaars and archaeological documentation centers.
unless some discoveries had been made there. With regard to the first state-
ment, it could be argued that if nothing had been found at Ziwiye in 1947 the
peasants might not want to reveal such information because they would lose
their jobs and their much-needed wages. Furthermore, one wonders why
they would voluntarily point out the alleged find spot, since if they actually
knew it they surely would want to protect it, to exploit it for themselves, and
not turn it over to strangers. The second statement could be countered by
noting that a dealer would dig at Ziwiye, or any other site, if he believed on
the basis of reports that something of value had been found there at some
time. Thus I suggest that the statements of the local villagers and the pres-
ence of a commercial dig at Ziwiye cannot by themselves be accepted as
sufficient evidence that something of importance was found at Ziwiye.
If we were cognizant of something concrete concerning the manner of
the transfer of the material from its alleged find spot to Teheran, then we
would perhaps be in a better position to feel more secure that something
may indeed have been found there in 1947. Did, for example, dealers pur-
chase the objects at the village of Ziwiye itself? Or did villagers known to
live at Ziwiye approach dealers in Sakkiz or Hamadan (Samedi 1960, 18), or
Teheran? But if Godard and Ghirshman had specific information with regard
to these matters, they did not offer it in print. And although I do not doubt
that they believed in 1950 that the objects they presented as from Ziwiye
actually came from there, I argue that they actually could not have known.
Thus, if one believes on the basis of the published information that
something was found at Ziwiye in 1947, it must be understood that such
belief is based ultimately on faith, on the acceptance of circumstantial and
hearsay evidence, and not on data objectively derived. This conclusion, it
must be stressed, does not deny the possibility, even the probability, that a
major find occurred at Ziwiye in 1947; it merely places the problem into a
perspective and defines its parameters.
If, then, one tentatively accepts the possibility/probability that some-
thing was found at Ziwiye in 1947, the next question follows inevitably:
what was found at that time? It is probable that most scholars would no
doubt answer by citing the objects published by Godard 1950, and it has
already been noted how those who expressed some doubts about attri-
bution, nevertheless accepted Godards authority. It might be argued that
surely the gold gorget, trapezoidal plaques, one or more other plaques and
strips, all stylistically related, must have come from Ziwiye.27 Wilkinson
27 Indeed, it might be argued that if the stylistically related objects did not come from
(1975, 7) summarizes this attitude: Godards description of the pillage and

later the mutilation of the finds, particularly those of gold, can be considered
the most authoritative. Some scholars would probably accept Godards sil-
ver and ivory objects as well. But immediately the question arises, which
silver objects? Could not the silver plaque illustrated in Godards fig. 109
have come from another site, one further north?28 And which ivory objects?
Could not the plaque illustrated in figs. 91 and 92 have come from another
site, especially because it is not stylistically related to the other published
ivories?
When every object published by Godard (not to mention Ghirshman) is
examined against this background, the same conclusion expressed above
forces itself upon us: the acceptance of any object published by Godard as
in fact deriving from Ziwiye can be defended on the basis of faith alone,
on the subjective (and I believe mistaken) belief that Godard really knew
what was found at Ziwiye. The acceptance of any of Godards objects must
be recognized for what it is, a subjective one; it should not be camouflaged
to appear as an objective, archaeologically verifiable conclusion. However,
as expressed above with regard to the previous question, this conclusion
does not categorically deny the possibility that Godards material derived
from Ziwiye. It merely states that from any historical or archaeological
point of view we do not know which particular objects were found there.
All/most/some/none? Here is the essence of the question.
Was anything found at Ziwiye after 1947? What can one say about the
provenience of the hundreds of objects from Ziwiye purchased from deal-
ers in Europe, Iran, and the United States beginning in the early 1950s and
continuing to the present? To my mind the most generous answer possible
Ziwiye they surely collectively came from another site, X. But this hypothesis only begs the
question: we still would not know where the site was located, whether or not in fact the
objects were found together, or whether they were found at one time. Personally, I accept
the possibility that they came from one find; but the acceptance is conditioned by the
reservations expressed in the text.
28 If the silver plaque had been correctly excavated at Ziwiye we would have in hand the
southernmost existing example of Urartian art in Iran. As it stands now, we merely have
a fine example of Urartian art that probably came from somewhere in Iran. For a state-
ment about two Urartian helmets in the British Museum that are said to come from Azer-
baijan and Luristan respectively, see my review of P. Calmeyers Reliefbronzen in babylonis-
chem Stil in a forthcoming issue of JAOS. And with regard to the citation of unexcavated
objects from unexcavated sites in historical reconstructions, see my comments concern-
ing the gold Achaemenid bracelet in the Karlsruhe Museum in Appendix B of Muscarella
1977.
is that the same conclusions presented above for the material published
by Godard and Ghirshman apply equally to the post-1950 material. But
these objects are many stages removed in time and place from the material
published in 1950 and as such offer even less authority (and require even
more faith) for acceptance into the corpus. In this context it will be recalled
that the first suspicious objects began to appear on the art market in the mid
1950s, after the publication of Godard 1950; these objects were attributed
to Ziwiye by the same vendors who had labelled, and continued to label,
other objects as from Ziwiye.
Is information volunteered by the Ziwiye villagers of any help in resolving
this question? Dyson (1957, 33; 1963, 34) reported that although commercial
digging continued at Ziwiye from 1947 to 1956, nothing of significance had
been found. This information came from the local villagers working for
the dealer licensed to dig there. It might be argued that this information
does derive from first-hand sources and indicates that in fact nothing of
importance was found at Ziwiye, at least after 1947, that all the material
said by dealers to have come from there after this date must have derived
from other sites. However, it could equally well be argued that perhaps
the workmen may have been under orders to suppress any information
regarding actual finds for several reasons, one of which could have been to
keep eager archaeologists away. Furthermore, it could be argued that the
enormous amount of material found in 1947 was hoarded by dealers and
released over the years as a sound business practice. In the final analysis,
statements by local villagers cannot be verified and have no objective value
as an argument either to support or to reject any theory concerned with
what was or was not found at Ziwiye after 1947. And any theories regarding
dealers inventories and sales techniques can ultimately only be based on
the claims of the dealers themselves, who, being parti pris, can hardly be
depended upon to solve our problem.
The answer to the final questionwhat historical and archaeological
value does the Ziwiye material have?has been anticipated by the con-
clusions already suggested. If it cannot be treated meaningfully as a unit
representing the archaeological finds from one site, it therefore must fol-
low that any historical or archaeological theories or conclusions based on
the existence of a collective treasure or hoard from Ziwiye will be of limited
value, if of any value at all. To treat all, or a subjectively selected group, of the
published objects as a unit-find is to misrepresent, and therefore to abuse,
the few facts available to us.
But the objects exist, and the genuine ones obviously have a significant
value. Indeed, not only are they exquisite works of art, they are of great
importance for the history of ancient Iranian art.29 As such, they must be
integrated into our study of Iranian and Near Eastern art in general. The only
caveat that should be ever present, indeed that should underlie any inclu-
sion of the Ziwiye objects into discussions of ancient art and culture, is
that they are to be considered as individual pieces, with no known prove-
nience, and without any firm evidence that they were found juxtaposed to
any other objects. They can tell us nothing about a hoard/burial deposited
at a particular place by a particular people; they surely can tell us nothing
historical about the site of Ziwiye. Historical and archaeological hypothe-
ses, for example those regarding the artistic and chronological relationship
of Iranian or Near Eastern art to Scythian art, can be maintained only with
this information in the forefront of the discussion. And, not parenthetically,
it will be obvious that the same constraints logically apply for the site of
Kaplantu and the objects said by dealers to have come from there.
The failure of scholars to critically examine the data and to question
them, the failure to recognize the situation for what it actually was, not for
what was claimed for it, have led to the creation of the Ziwiye problem.
This problem laid out before us forms a paradigm of a larger problem of
Iranian archaeology. For if one examines the corpus of unexcavated mate-
rial claimed by dealers and then published by scholars as deriving from
Hamadan, Marlik, Luristan, Amlash, etc., one soon realizes that there is
a Marlik and a Marlik, a Hamadan and a Hamadan, and so forth. The
objects said to come from these sites are sold by the same dealers, the sto-
ries concerning their excavation are told by the same story tellers to the
same receptive audience, an audience of credulous individuals whose desire
for objects qua objects is so insatiable, they will destroy their own discipline
in order to acquire and publish them.
Acknowledgements
I wish to thank Professors Louis D. Levine, Irene Winter and T. Cuyler

Young, Jr. for reading this paper in manuscript form and for their helpful
suggestions and comments. I alone assume responsibility for the papers
nature and conclusions and for not always following my colleagues advice.
29 I do believe that the available information allows us to conclude that the objects derived
from some site(s) within Iran.

Bibliography
The Bibliography is arranged chronologically rather than alphabetically in

order to document the publications and discussions of the Ziwiye material in
the historical manner in which they were presented.
1948
Godard, Andr. Comments in Muse Cernuschi, Iran: Pices du Muse de Teheran, du
Muse du Louvre, et de Collections Particulires (Paris) 9, 14.
1949
Godard. Andr. Le Trsor de Ziwiy, Compte Rendus de LAcadmie des Inscriptions
et Belles Lettres, 168172.
1950
Dimand, Maurice S. Treasures of Iranian Art, BMMA 8, 5, 145146.
Godard, Yedda. a) Histoire dun Trsor, France Illustration, April 8, 331333.
b) The Arabian Nights Story of the Gold Treasure of Ziwiye, ILN, May 8, 714
715.
Godard, Andr. Le Trsor de Ziwiye (Haarlem).
Ghirshman, Roman. Le Trsor de Sakkez, les Origins de lArt Mde et des Bronzes
du Luristan, Artibus Asiae 13, 181206.
1951
Crousset, Ren. Dernires Vues sur lArt des Steppes, La Revue des Arts 2, 107110.
Dussaud, Ren. Revue of Godard 1950, Syria 28, 287289.
Revue of Ghirshman 1950, Syria 28, 289291.
Godard, Andr. Propos du Trsor de Ziwiy, Artibus Asiae 14, 240245.
Wiedner, E.F. Note in AFO XV, 151152.
1952
Falkner, Margarete. Der Schatz von Ziwije, AfO XVI, 129132.
Wilkinson, C.K. Some New Contacts with Nimrud and Assyria, BMMA 10, 8, 233
240.
1954
Ghirshman, Roman. Iran from the Earliest Times to the Islamic Conquest (Penguin).
Schefold, K. Die Iranische Kunst des Pontuslnder, in Handbuch des Alterwis-
senschaft (Munich) 423453.
Sulimirski, T. Scythian Antiquities in Western Asia, Artibus Asiae XVII, 282318.
vanden Berghe, L. Iran, De Stand van de Archaeologische Onderzoekingen VIII
De Kurdistan, Jaarbericht Ex Orient Lux 13, 382384.
1955
Anonymous. Assyrian and Persian Gold and Silver: Masterpieces from Recent Dis-
coveries in the Metropolitan Museum of Art, New York, ILN, April 16, 699.
Frankfort, Henri. The Art and Architecture of the Ancient Near East (Penguin).
Wilkinson, C.K. Assyrian and Persian Art, BMMA 13, 8, 213222.
1956
Anonymous. New Treasures of Teheran Museum: Achaemenian Gold and Later
Silver, ILN, July 21, 107.
Barnett, R.D. The Treasure of Ziwiye, Iraq XVIII, 111116.
Bussagli, M. Objects from the Ziwiyeh Treasure, in Mostra dArte Iranica (Milan)
119124.
von der Osten, H.H. Die Welt der Perser (Stuttgart).
Wilkinson, C.K. Two Ancient Silver Vessels, BMMA 15, 1, 915.
1957
Carter, Dagny. The Symbol of the Beast (New York).
Dyson, R.H., Jr. Iran 1956, UPMB 21, 1, 2739.
Kantor, Helene, J. Goldworks and Ornaments from Iran, Cincinnati Art Museum
Bulletin 5, 2, 920.
Needler, Winifred. Four Near Eastern Antiquities Lent by Mr. Joseph H. Hirsch-
horn, ROM Bulletin, 25, 711.
1958
Amandry, Pierre. Orfverie achmenide, Antike Kunst 1, 923.
1959
vanden Berghe, L. Archologie de lIran Ancien (Leiden), 111115.
1960
Benson, Jack L. Unpublished Griffin Protomes in American Private Collections,
Antike Kunst 2, 5870.
Ghirshman, Roman. Rich Treasures of Persian Animal ArtRecently Discovered,
ILN, April 4, 550551.
Kantor, Helene J. A Fragment of a Gold Appliqu from Ziwiye and Some Remarks
on the Artistic Traditions of Armenia and Iran during the Early First Millennium
B.C., JNES 19, 1, 14.
Samedi, H. Les Dcouverts Fortuites (Teheran).
Wilkinson, C.K. More Details on Ziwiye, Iraq XXII, 213220.
1961
Ghirshman, Roman. Le Trsor de Ziwiye, in 7000 Ans dArt en Iran (Paris) 81102.
Hoffmann, Herbert. The Persian Origin of Attic Rhyta, Antike Kunst 4, 1, 2126.
Melikian, Souren. Les Fouilles clandestines de lIran, Connaissance des Arts 117,
November, 6673.
Parrot, Andr. The Arts of Assyria (New York) 138146.
1962
Barnett. R.D. Median Art, Iranica Antiqua II, 1, 7795.
Ghirshman, Roman. a) Le Rhyton en Iran, Artibus Asiae XXV, 5780.
b) Der Schatz von Ziwije, in Kunstschtze aus Iran (Zurich) pp. 7186; see
also same entry in 7000 Jahre Kunst in Iran (Essen) 93114.
Godard, Andr. LArt de lIran (Paris).
Tuchelt, K. Tiergefsse in Kopf- und Protomengestalt (Berlin).
van Ufford. L. Byvanck-Quarles. Le Trsor de Ziwiy, BABesch 2539.
1963
Dyson, R.H. Jr. Archaeological Scrap; Glimpses of History at Ziwiye, Expedition 5,
3, 3237.
Potratz, J. Die Skythen in Sdrussland (Basle).
van Ufford, L. Byvanck-Quarles. Propos du Trsor de Ziwiy, BABesch 100105.
Wilkinson, C.K. Treasure from the Mannaean Land, BMMA 21, 5, 274284.
1964
Boehmer, R.M. Volkstum und Stdte der Manner, Baghdader Mitteilungen 3, 11
24.
Goldman, Bernard. Early Iranian Art in the Cincinnati Art Museum, Art Quarterly
27, 324341.
Ghirshman, Roman. a) The Arts of Ancient Iran (New York) 98125.
b) Invasions des Nomads, in Dark Ages and Nomads, ed. Machteld Mellink
(Istanbul 1964) 38.
Porada, Edith. The Treasure of Ziwiye, in 7000 Years of Iranian Art (Smithsonian
Institute) 2527, 8586.
1965
Amandry, Pierre. a) LArt Scythe archaique, AA 891913.
b) Un Motif Scythe en Iran et en Grce, JNES 24, 3, 149160.
Culican, William. The Medes and Persians (New York).
Dyson, Robert H., Jr. Protohistoric Iran as Seen from Hasanlu, JNES 24, 3, 205207.
Huot, Jean-Louis. Persia I (Archaeologia Mundi, World Publishing Co., New York)
139140.
Muscarella, Oscar White. A Fibula from Hasanlu, AJA 69, 233240.
Porada, Edith. The Treasure of Ziwiye, in The Art of Ancient Iran (New York) 123
136.
Terrace, E.L.B. The Ziwiye Treasure, in BullMFA LXIII, 1315.
Young, T.C. Jr. A Comparative Ceramic Chronology for Western Iran, 1500500B.C.,
Iran III, 5385.
1966
Amandry, Pierre. A Propos du Trsor du Ziwiy, Iranica Antiqua, VI, 109129.
Muscarella, Oscar White. Review of Porada 1965 in AJA 70, 3, 380381.
Shepherd, Dorothy. Four Early Silver Objects from Iran, Bulletin of the Cleveland
Museum of Art LIII, 2, 3849.
Terrace, E.L.B. Ancient Near Eastern Art, in The Pomerance Collection of Ancient Art
(Brooklyn) 4248.
Muse Rath. Trsors de lAncien Iran.
van Loon, M. Urartian Art (Istanbul).
1967
Jettmar, Karl. Art of the Steppes (Baden-Baden).
van Loon, M. Review of M. Mellink (ed.) Dark Ages and Nomads c. 1000 B.C. (Istanbul,
1964) in Bibliotheca Orientalis XXIV, 1/2, 2126.
Wilkinson, C.K. Two Ram-Headed Vessels from Iran (Bern).
1968
Azarpay, G. Urartian Art and Artifacts (Univ. of California).
1969
Amiet, Pierre. Notes dArchologie Iranienne, La Revue du Louvre 19, 6, 325338.
1970
Farkas, Ann. The Near East, in Animal Style Art (Asia House, New York) 1927.
Lorenz, Eva. RaubgrbereiNicht Aktenkundig, Antike Welt 1, 3543.
van Loon, M. Review of Potratz 1963 in JNES 29, 6672.
1971
Hrouda, B. Vorderasien I (Munich).
Moorey, P.R.S. A Catalogue of the Persian Bronzes in the Ashmolean Museum (Ox-
ford).
Muscarella, Oscar White. Hasanlu in the Ninth Century bc and its Relations with
other Centers of the Near East, AJA 75, 263266.
1972
Burney, Charles, Lang, D.M. The Peoples of The Hills (New York).
Phillips, E.D. The Scythian Domination in Western Asia , WA 129138.
1973
Barnett, R.D. A Review of Acquisitions 19631970 of Western Asiatic Antiquities (2),
BMQ 37, 34, 119137.
Calmeyer, Peter. Reliefbronzen in babylonischem Stil (Munich).
Ghirshman, Roman. A Propos du Trsor de Ziwiy, JNES 32, 4, 445452.
Muscarella, Oscar White. Excavations at Agrab Tepe, Iran, MMAJour 8, 4776.
1974
Ghirshman, Roman. Une Mde sur les Bas-Reliefs de Nimrud, Iraq 36, 1/2, 3738.
Mallowan, Max and G. Herrmann. Ivories from Nimrud Fascicle III Furniture from
SW.7 Fort Shalmaneser (London) 5557.
Muscarella, Oscar White. The Iron Age at Dinkha Tepe, Iran, MMAJour 9, 3590.
1975
Orthmann, W. Die Kunst des Iran, in Der Alte Orient, Propylen Kunstgeschichte,
93.
Porada, Edith. Iranische Kunst, in Ibid. 363398.
Wilkinson, C.K. Ivories from Ziwiye (Bern).
1976
K. Ishiguro. The Mr. and Mrs. Ishiguro Collection of Ancient Art I (Tokyo).
Muscarella, Oscar White. Review of Mallowan, Herrmann 1974 in JNES 35, 3, 208210.
1977
Muscarella, Oscar White. Unexcavated Objects and Ancient Near Eastern Art, in
press.
chapter thirty-eight
MEDIAN ART AND MEDIZING SCHOLARSHIP*
More than two decades have passed since Gza de Francovitch wrote that
Median art is a myth, that not a single work exists of proven Median
origin, and, further, that its discussion involves one in the realm of myth
and fable .1 That this strongly presented view still obtains, is still forcibly
valid, even more so after the excavation of two apparent Median sites in
western Iran, and after the appearance of more scholarly writings that claim
to have unraveled and perceived the characteristics of Median art, will be the
conclusion of this paper.
The political and social history of the Medes is difficult to chart and
reconstruct in any of its specifics and at any point in time, and here we
can only deal with it in the broadest terms.2 The sources are primarily the
contemporary, often terse references in the Assyrian annals and records that
begin in the ninth century bc (836 bc) and continue into the seventh, the
brief but significant references in the seventh-sixth century sections of the
Babylonian Chronicle, and the later fifth-century bc writings of Herodotus
(l. 95106). The Assyrian annals indicate that the Medes were powerful and
occupied many towns and cities, each under a chieftain, they were attacked
periodically by the Assyrians, and they fought with their neighbors. By the
last two or three decades of the seventh century, some political events
unknown to us had occurred, for by this time the Medes (or some Medes)
became united under one king. This unification strengthened them to the
degree that in the years between 614 and 609 bc under Cyaxares, the king of
the Medes, they invaded Mesopotamia and participated in the destruction
of the Assyrian state. By 585bc it seems that Cyaxares had secured control
* This article originally appeared as Median Art and Medizing Scholarship, Journal of
Near Eastern Studies 46, no. 2 (1987): 109127.

I wish to thank Susan Patullo for her comments on various aspects of this paper.
1 Gza de Francovitch, Problems of Achaemenid Architecture, East and West 16 (1966):
257.
2 I present here only the barest outline. Recent important research makes it clear that we
know far less about Median history than hitherto assumed; see, in particular, Payton R. Helm,
Herodotus Medikos Logos and Median History, Iran 19 (1981): 8590, with bibliography.
1000 chapter thirty-eight
over most of the peoples of western Iran including the Persians, and he
or his son Astyages had advanced to the Halys river in Anatolia, where he
maintained a border with the powerful Lydians. Within a generation all
was lost. Cyrus II, the king of the Persians, rebelled against his father-in-law
Astyages and in turn subjugated the Medes (553550bc). With this event
began a new and, from a modern perspective, decidedly clearer phase of
Iranian and Near Eastern history.
The heartland of the Medes in western Iran centered around Hamadan
(Ecbatana) and extended for some distance to the southwest, northeast, and
east in the Alvand mountain range.3 The Assyrian annals record wealthy
Median cities, but the only kinds of booty mentioned are weapons, cattle,
donkeys, horses, camels, and, rarely, lapis lazuli,4 the latter without doubt
acquired by Median trade further east. By the time of their unification or
shortly thereafter, they seem to have acquired the means to supply them-
selves with more substantial wealth. This may be inferred from the sixth-
century bc section of the Babylonian Chronicle (7, ii, 34)5 that records
Cyruss defeat of Astyages: Cyrus (marched) to Ecbatana, the royal city. The
silver, gold, goods, property which he carried off as booty (from) Ecbatana,
he took to Anshan. The Chronicle does not define the nature and char-
acteristics of the precious metals or of the goods and property, except to
record that portable material is being discussed. Thus, whether the gold and
silver was in the form of bullion or crafted artifacts, and whether Median
Kleinkunst or state or religious objects were included in the spoils, is not
vouchsafed. But something valuable was taken by Cyrus from Ecbatana, and
I do not believe one can exclude the possibility that these valuables included
products of Median craftsmen.
There is yet another possible interpretation that may be brought forth
regarding the nature of the spoils taken from Ecbatana, an interpretation
3 A number of scholars have argued that the Median area of control extended as far
northeast as Tehran on the assumption that the Mt. Bikni associated with the Medes in the
Assyrian records is Mt. Damavand. However, L.D. Levine, Geographical Studies in the Neo-
Assyrian Zagros-II, Iran 12 (1974): 118 f., nn. 167, 168, following Knig, correctly demonstrated
that Mt. Bikni is most probably a mountain in the Alvand range, near Hamadan. J. Reade,
Kassites and Assyrians in Iran, Iran 16 (1978): 38, continues to support the Bikni-Damavand
connection.
4 D.D. Luckenbill, Ancient Records of Assyria and Babylonia (Chicago, 1926), vol. 1, pars.
795, 812; vol. 2, pars. 540, 566.

5 For the references to the Babylonian Chronicle mentioned here and elsewhere in the
text, see A.K. Grayson, Assyrian and Babylonian Chronicles (Locust Valley, New York, 1975),
pp. 91, 106.
median art and medizing scholarship 1001
suggested to me by T. Cuyler Young, Jr.: the spoils could have been booty
that the Medes themselves took from the Assyrians more than a half-century
earlier. Explicitly argued in this view (assuming it means that only Assyrian
valuables were kept at Ecbatana) is the conclusion that, for whatever rea-
sons, the Medes chose not to or could not employ their own craftsmen to
create works. Indeed, this conclusion is one that could follow after the fact,
so to speak, after reviewing the evidence presented in the main body of this
paper. But that evidence allows one only to recognize what does or does not
exist as a result of modern discovery, not what may have actually existed in
the sixth century bc. I am of course aware of the limitations in the use of
the Babylonian Chronicle to prove the existence of Median art and artifacts,
whether we posit them to be mundane objects or sumptuous products of a
Staatskunst (infra), just as I am aware of its limitations for disproving such
existence. Here I am merely exploring the range of possible interpretations,
rather than attempting to make definitive statements.
That the gold and silver recorded in the Babylonian Chronicle may not
have been solely in the form of bullion or Assyrian booty and may have
included objects crafted by Median artisans is perhaps suggested, albeit
tenuously, by the later inscription of Darius at Susa (DSf). Here we read that
the goldsmiths who wrought the gold, those were Medes and Egyptians;
the very same people are also said to be the ones who adorned the walls.6
Whatever the meaning of adorned,7 the Medes were working together with
Egyptians as goldsmiths, which suggests, even if it does not prove, that some
tradition of Median gold working existed prior to the early fifth century bc.
Whether this tradition existed in pre-Cyrus times or began during his or
Dariuss reign is not vouchsafed, butagain, perhapsthe evidence of the
Ecbatana spoils would support the former position.8
Given then both the historical background of a powerful, united Median
state in the late seventhearly sixth century bc and the references to silver
and gold objects, scholars have posited that a Median art and iconography
probably existed by the late seventh century and perhaps earlier, and that
objects of sumptuous and courtly quality, inter alia, were crafted by Medes.
To give one example of the thought process concerning what might be
6 R.G. Kent, Old Persian (New Haven, 1950), p. 144.

7 M.C. Root, The King and Kingship in Achaemenid Art (Leiden, 1979), pp. 7ff.
8 Root (ibid., p. 32, n. 95) claims that Median craftsmen worked at Nebuchadezzars court,
but from the reference she supplies, it is far from certain that the Medes in Babylon were
craftsmen; the word ba-ak-tu is not a craft designation (Ira Spar, private communication).
expected of the Medes, M. Root states, It is not unlikely that at important

centers such as Ecbatana buildings were adorned with sculptures in stone
or glazed brick relief. Also, a tradition of monumental representational art
may well have been maintained such as wall painting and wooden relief.9
This observation is viable, as are those postulating the existence of portable
Median art, Kleinkunst, as well as sumptuous works. Such assumptions
not only do not conflict with, but are based on our perceptions of the
interrelationships of ancient Near Eastern art and politics. In this context,
scholars might also anticipate a Median writing system for economic and
political transactions and major architecture to accommodate the royal
family as well as local officials.
Before proceeding further, brief, but pertinent, comments and a caveat
on the terms Mede and Median art are called for. A people called Medes,
distinct from other peoples in western Iran, are known in Iran from the
ninth century bc. By the last two decades of the seventh century there is
a Median army under a king, the implication of this being that a cen-
tral Median authority is present. Is it legitimate to assume that by the later
seventh century the Median kingdom incorporated other peoples in west-
ern Iran who were thereby considered to be Medes in a political sense and
who were components of the Median army? The Babylonian Chronicle (3.5)
records that the Manneans in 616 bc were still a distinct political entity, inas-
much as they were allies of the Assyrians against the Babylonians, which
signifies that not all western Iranians (meaning here peoples living in west-
ern Iran regardless of languages spoken) were included in the Median state
at this time. But what of the Manneans a few years later, and what of oth-
ers, Harhar, Missi, Ellipi, etc., who cease to be mentioned in the late Assyrian
texts? Was there a Media before and a Greater Media after the last decade or
two of the seventh century bc? Would Median material culture and iconog-
raphy in the earlier period basically reflect one form, purely Median, and
in the later another composed of several cultures? And if so, what part was
ethnically Median and what parts Ellipian, Mannaean, Missian, etc.?
Indeed, only Media is mentioned in 550bc when Cyrus defeats Astyages, but
what is Media? We do not know the answers to all the questions posed. As
we have no conception concerning the artistic and iconographic nature of
most of the peoples and political entities of western Iran (excepting Luris-
tan) either before or after the late seventh century, any attempt to isolate
Median art must take into account both the complex cultural and ethnic
9 Ibid., pp. 32 f.
background of the area and the question of precisely at what time and where
in Iran our attention is being focused.10
Now, having in mind from the above discussion an anticipation of what
might be reflected in the archaeological record with regard to the culture
of the Medes, what must have existed, what in reality do we have? First
of all there is no Median writing extant in Median or in Elamite script,
not a single sign on any object, artifact, or work of art.11 With regard to
architecture an apparently different situation exists. In the mid-1960s two
almost certainly Median sites were excavated, Godin Tepe (Period II) and
Nush-i Jan, southwest and south of Hamadan respectively.12 Both sites were
founded sometime in the late eighth century bc, that is, at a time before
the unification under one king, presumably Cyaxares. They both have a
large building with a multi-columned hall, a form known earlier at Hasanlu
in northwestern Iran and later at Pasargadae and Persepolis; Nush-i Jan
has, in addition, a fort and two fire temples, all remarkably well preserved.
Ecbatana remains unexcavated,13 hence these two sites will serve as the focus
of future research on Median architectural features and characteristics.
10 See D. Stronach, Excavations at Nush-i Jan, 1967, Iran 7 (1969): 4, and Helm, Herodo-
tus, p. 87. Some of the thoughts expressed here have been sharpened by conversations with
T. Cuyler Young, Jr. See also my brief cautionary comments in Journal of the Metropolitan
Museum of Art (MMJ) I (1968): 17 f.
11 Thus, John Curtis, Nush-i Jan III: The Small Finds (London, 1984), p. 14, correctly refrains
from identifying as Median a few broken cuneiform signs incised on a silver fragment (pl. 8,
100) from Nush-i Jan. The absence of Median writing, of course, does not mean that it
never existed; see I.M. Diakonov apud R. Ghirshman in Bibliotheca Orientalis 15, no. 6 (1958);
260. The pre-Achaemenian Persians used Elamite for their inscriptions, as is evidenced by
seatings from Persepolis with the inscription, Cyrus of Anshan, son of Teispes.
12 T. Cuyler Young, Jr., Excavations at Godin Tepe: First Progress Report (Toronto, 1967);
T. Cuyler Young, Jr. and L.D. Levine, Excavations at Godin Tepe: Second Progress Report (To-
ronto, 1974). D. Stronachs reports on Nush-i Jan have been published in Iran 1969, 1971, 1973,
1974, 1975, 1978; see also Curtis, Nush-i Jan III. W. Kleiss, Zwei Pltze des 6. Jahrhunderts
v. Chr. in iranisch Azerbaidjan, Archologische Mitteilungen aus Iran (AMI) 9 (1976): 114ff.,
emphatically considers a surveyed site, Chorbulag, northeast of Maku in northwestern Iran,
with sixth-century bc pottery, to be der Anlage um einen medischen Herrensitz (p. 114)
and das einzige Gebude im Nordteil von Iranisch-Azerbaidjan, das von den Medern
errichtet worden ist (p. 116). Yet, while the site may date to the Median period in general
terms, it may not be called Median in the narrow, ethnic sense. There is no evidence that
Baba Jan (C. Goff, Excavations at Baba Jan: The Pottery and Metal from Levels III and II,
Iran 16 [1978]: 41 f.) is Median.
13 A few sondages have been made there. A large number of unexcavated objects, some
ancient, some modern forgeries (infra), have been assigned to Hamadan by dealers and
scholars with the assumption, implicitly or explicitly expressed, that they are Median; see
my Excavated and Unexcavated Achaemenian Art, in D. Schmandt-Besserat, ed., Ancient
Persia: The Art of an Empire (Malibu, 1980), pp. 31 ff.
It must be noted and stressed that it is precisely the coincidence of their

chronological position and geographical location that permits archaeolo-
gists to designate the two sites as Median with little reservation, for if they
had been uncovered elsewhere in Iran, they could have been assigned to dif-
ferent peoples. But whereas their massive building remains inform us of the
existence of Median architects, of their building techniques and plans and
their secular and religious structures, both sites yielded not a single arti-
fact (except pottery) or motif that distinguishes itself as Median, let alone
as Median art or iconography. This lack may be explained by the possibly
historically significant fact that both sites were abandoned, and little aside
from scraps and pottery was left behind. Whatever the reasons, there is no
Median art represented at the only two assumed Median sites excavated to
date (see infra).
Median art, objects implicitly or explicitly assumed to have been made by
Medes, has been identified by scholars among the precious material found
or allegedly found beside the Oxus River, among finds from the Caucasus
and Soviet Union and among objects or elements in the so-called Ziwiye
treasure. Also considered to be examples of Median art are a number of stray
unexcavated objects, some ancient, others not, which are claimed by dealers
and cooperating scholars to have derived from western Iran. A series of rock-
cut tombs in Iran and Iraq and a few objects actually excavated in western
Iran complete the corpus of putative Median art. To these putative Median
creations we now turn for review.
When O.M. Dalton published (1926) the British Museums collection of

material claimed to have been found at a site on the Oxus River in 1877,
he singled out among the manifestly Achaemenian objects two that he
believed appear to be older than the time of the Persians, and which
may have a Median origin or which provisionally may be ascribed to
that Median art of which so little is known 14 Considered to be Median
14 O.M. Dalton, The Treasure of the Oxus (1926; London, 1964), nn. 2, 10: in the first,
1905, publication, only the sheath was singled out, pp. 18, xvi, xxii. For my comments on the
provenience of the treasure, see my 1980 paper, Excavated and Unexcavated Achaemenian
Art, p. 26. I wish to make two modifications to that paper: it was R.D. Barnett, not Dalton,
who wrote the preface to the 1964 edition quoted on p. 27; and it should have been made
clearer that the silver rhyton published by Dalton, lx (not XL), no. 178, was not claimed to be
part of the Oxus treasure (p. 30) but, rather, from an alleged separate find at Erzerum.
In two recent articles by B.A. Litvinskiy and I.R. Pichikyan, The Temple of the Oxus, JRAS
1981, pp. 134 ff.; and Monuments of Art from the Sanctuary of Oxus, Acta Antiqua 27 (1983):
25 ff., it is claimed as an archaeological fact (indisputable) that the site of Takht-i Sangin
were a gold bowl (no. 18) decorated with six pairs of rampant winged lions
separated by a lobe, and a gold akinakes sheath (no. 22) decorated with
a scene, repeated five times, of a horseman shooting arrows at a rampant
lion. Criteria used to establish the tentative Median attribution were the
stylization of the lions on the bowl, the apparent Assyrian influences of
the hunt scene, and the high tiara headdress on the horseman. In 1921
E. Herzfeld accepted the sheath as medische Arbeit, and in 1941 he added
the gold plaques from the Oxus group.15
R.D. Barnett in 1957, 1962, and 1968 transformed Daltons tentative attri-
bution into a certainty.16 Discussing only the sheath, he correctly compared
it in form and scalloped border decoration to the example represented on
the well-known Audience Hall (Apadana) relief, where it is worn by a Mede.
Because of its presence on this relief, the akinakes sheath is a traditional
Median type of scabbard, and the scalloped border on the Oxus sheath
betrays it as Median,17 Having thus concluded that the akinakes is peculiar
to Medes, Barnett then leaped from a culturally-ethnically defined object
to a chronological position, automatically perceiving the Oxus sheath as
Median in a pre-Achaemenian sense, a product of the seventh century bc.
And so positive was he of this attribution that he claimed forcefully and
without qualification that the horseman on the Oxus sheath can only be
Astyages himself.
on the right bank of the Oxus River is the site from which the Oxus treasure derived. Other
than nineteenth-century opinions, the sole piece of archaeological information offered to
support this claim is the discovery at the site of an extraordinary ivory akinakes sheath (JRAS,
1981, pl. 1; Acta Antiqua, 1983, fig. 2) decorated in relief with a rampant lion clutching a stag.
Because this sheath is the same size as the gold example from the treasure (26.7cm) and,
to the authors, of the same date, the ivory find tends to confirm the traditional opinion
about the find spot of the Oxus Treasure . But on their own merits alone, the date of
the sheath, found in a third-century bc temple with third-century bc material, as well as its
find-spot cannot possibly support the fiat that the original site of the Oxus treasure is now
revealed. I refuse to accept the claim and find its absoluteness not proper in archaeological
discourse.
15 Dalton, Treasure of the Oxus, pls. 14 and 15; E. Herzfeld, Kattische und khaldische
Bronzen, in K. Regling and H. Reich, eds., Festschrift zu C.F. Lehmann-Haupts sechzigsten

Geburtstage (Vienna and Leipzig, 1921), p. 154; idem, Iran in the Ancient East (London, 1941),
pp. 205, 267 f. See also H.H. von der Osten, Die Welt der Perser (Stuttgart, 1956), p. 54, pl. 73, for
other items added to the list of Median objects in the treasure.
16 R.D. Barnett, Persepolis, Iraq 19 (1957): 76; idem, Median Art, Iran. Antiqua 2 (1962):
7795, esp. 7880; idem, The Art of Bactria and the Treasure of the Oxus, Iran. Antiqua 8
(1968): 36, 38.
17 Idem, Median Art, pp. 79 f. For the Persepolis relief sheath, see E. Schmidt, Persepolis,
vol. I (Chicago, 1953), pl. 120.

Over the following years E. Porada, W. Culican, M. van Loon, E. Akurgal,

and A. Farkas also considered the Oxus sheath to be Median, both cul-
turally and chronologically.18 Among them, van Loon further discussed the
Oxus gold bowl, which, following Dalton, he considered to be Median; and
he identified the horseman on the sheath not as Astyages, but probably
Cyaxares, thereby dating the sheath still earlier in the Median period than
did Barnett.
R. Ghirshman not only vigorously supported Herzfelds and Barnetts con-
clusions concerning the Oxus sheath and gold plaques,19 but expanded the
perception of Median art and culture by presenting as parallels to the gold
plaques some unexcavated bronze plaques that he believed derived from
Luristan and which, in his view of archaeology, culturally connected the two
distant areas. To Ghirshman, the art of Luristan is obviously Cimmerian; that
it is also in part Median is revealed by the plaques and by a bronze spouted
vessel with rear applique excavated at Hamadan (infra) that compares with
others he believed were excavated in Luristan. By some unexplained pro-
cess, these occurrences result in une symbiose mdo-cimmrienne, with
both peoples sharing religious views and cultural identities. Whatever this
means, it will be noted below that when discussing Median art and its char-
acteristics as manifested at Ziwiye, Ghirshman forgot or chose not to men-
tion Luristan; at Ziwiye a different symbiosis is revealed.
Barnetts writings specifically demonstrate that it is the akinakes sheath
form itself that generated the impulse to perceive the Oxus example not
only as a characteristic Median weapon but one necessarily made during
the Median period. A disinterested investigation of the evidence, however,
indicates a broader use and a longer life for this type of object than rec-
ognized. In the first place, on the Persepolis reliefs the akinakes sheath is
worn by Medes (assuming that all those figures wearing trousers and the
akinakes are in fact Medes),20 Medes, let it be stressed, living in the fifth
century bc. But there is more to record about the ethnic background of
18 E. Porada, The Art of Ancient Iran (New York, 1965); p. 140; W. Culican, The Medes and
Persians (New York, 1965), p. 140; M. van Loon, Urartfan Art (Iistanbul, 1966), p. 178; E. Akurgal,
Urartische und altiranische Kunstzentren (Ankara, 1968), p. 124; A. Farkas, The Horse and
Rider in Achaemenid Art, Persica 4 (1969): 69 ff.
19 R. Ghirshman, Le Trsor de l Oxus, les bronzes du Luristan et lart mde, in Vorderasi-
atische Archologie (Moortgat Festschrift) (Berlin, 1964), pp. 8894, esp, 8890; idem, The Arts
of Ancient Iran (New York, 1964), pp. 9094.
20 Root, King and Kingship, p. 98, n. 169; see also pp. 277; 279, n. 144; 281f.; and further
comments below.
those who wore the akinakes. Aside from the Medes (Delegation I), on the
Apadana reliefs they are worn or carried by Scythians (Delegation XI) and by
Delegation XVII, who are either Scythians, Sogdians, or Chorasmians. Fur-
ther, on the tomb reliefs they are worn by thirteen Vlker in addition to the
Medes.21 In short, the akinakes is not and cannot be associated solely with
Medes (see also notes 14, 23, and the Kelermes and Litoy-Melgunov sheaths,
below).
Of course Barnett was right to introduce the Audience Hall relief sheath
as an appropriate parallel to the Oxus sheath, but he was in error when he
misread its chronological message. The scalloped border and animal frieze
scene on the body of the relief sheath are explicit and formal parallels,
parallels that indicate an Achaemenian attribution for the Oxus sheath.
The hunt scene on the latter is Achaemenian in style and spirit and occurs
on a number of Achaemenian period seals where royal and apparently
other figures are depicted hunting animals from horseback, on camels, from
chariots, and on foot.22 And the high headdress worn by the horseman,
as Moorey has noted in his argument for an Achaemenian date for the
sheath, may be the turban [sic] worn by a number of the peoples within
the Achaemenid Empire .23 (The headdress seems not to be a turban, but
the general idea still obtains.) To identify, and with certainty, the horseman
as a specific Median king, of whom we know little historically and nothing
art-historically, even assuming the sheath to be pre-Achaemenian in date,
transgresses the boundaries of archaeological activity, all the more so when
we recognize the Achaemenian, probably fifth century bc, depiction before
us. And consequent to the fact that all horseman in the Achaemenian period
21 For identification of the Apadana relief figures and the tomb bearers, see G. Walser,
Die Vlkerschaften auf den Reliefs von Persepolis (Berlin, 1966), pp. 53ff., Falttafel I, and my
comments in my review of Walser in JNES 28 (1969): 283 f. See also B. Hrouda, Vorderasien,
vol. I (Munich, 1971), p. 286; and P.R.S. Moorey, Cemeteries of the First Millennium bc at Deve
Hyk, BAR International Series 87 (London, 1980), p. 55.
22 Viz., H. Frankfort, Cylinder Seals (London, 1939), pp. 221f., pl. 37d, f, h, i, m, n. See also
the silver disc in the Oxus treasure in Dalton, Treasure of the Oxus, p. 13, no. 24.
23 Moorey, Cemeteries, p. 57. In this context it is worth noting the silver rhyton from
Erebuni that depicts a heavy-set man riding a horse. The man bears a typical akinakes and
wears a tapering high headdress: B.N. Arakelian, Treasure of Silver Artifacts from Erebuni,
Sovietskaya Archeologica I (1971): figs. 14, dated to the fifth-fourth centuries bc. The man is by
no means necessarily a Mede. P. Calmeyer Zu einigen vernachlssigten Aspekten medischer
Kunst, in Proceedings of the IInd Annual Symposium of Archaeological Research in Iran, Iran
Bastan Museum, OctoberNovember, 1973 (Tehran, 1974), pp. 112, 118, n. 5, dates the Oxus sheath
to the second half of the fifth century bc on the basis of comparisons with Greek art; Farkas
Horse and Rider in Achaemenid Art, p. 70, also raised this issue but did not develop it further.
wore trousers, whether Medes, Persians, or others, it would be presump-

tuous to attempt an identification of the ethnic background of the Oxus
horseman.24
The Oxus gold bowl is less easy to attribute primarily because of the
relative crudeness of execution of the heads and front feet of the lions.
However, the bowls lobed decoration as well as the repousse technique
are typically Achaemenian,25 likewise, the classic Achaemenian circle and
comma motifs depicted on the lions haunches. Provincial is a convenient
clich, I admit, but I cannot see how one can exclude this bowl from the
Achaemenian corpus merely because it is crude, just as one cannot include
it as Median because it is crude.26 Finally, Herzfeld placed the Oxus gold
plaques in the Median period because of the dress of some of the depicted
figures and their relationship to figures represented on rock-cut tombs,
which he considered to be pre-Achaemenian. We shall return to these tombs
below.
Another object that Barnett discovered to be Median is the gold dag-
ger from Chertomlyk in the Crimea. Attribution is revealed in this instance
first by analogy: the hunt scene across the hilt connects it to the Median
Oxus sheath and secondly because the addorsed calf heads at the hilts apex
are a typical Urartian motif, a crucial ingredient in Barnetts Median for-
mula (infra).27 Indeed, the Oxus hunt scene parallel is obvious and pertinent;
24 See n. 20, above. On p. 281, n. 151, Root says that there are no known Achaemenian
representations of the king wearing trousers, but she notes that Darius III wears them on
the Alexander Mosaic; the Oxus sheath scene does not alter this observation. O. Kimball
Armayor, Herodotus Catalogue of the Persian Empire , Transactions of the American
Philological Association 108 (1978): 5, takes a rigid view of trousers contra Herodotus (1.71, 135;
3.87; 5.49; 7.61, 62), arguing that they were never worn by Persians, on which view see Root.
There can be little doubt that in battle (as Herodotus stated) and surely in the hunt, Persians,
and other people, wore trousers.
25 See, for example, a sixth-fifth century Archaemenian silver bowl excavated at Ushak
in western Anatolia (unpublished, in the Ankara Museum) and an example of the same date
(and possibly same site) in the Metropolitan Museum of Art, which D. von Bothmer, A Greek
and Roman Treasury, Bulletin of the Metropolitan Museum of Art (EMMA) 42 (1984): 25, no. 18
and also 19, incorrectly labeled Greek! Other examples include a silver vessel from Kazbek;
see A.M. Tallgren in Eurasia Septentrionalis Antiqua 5 (1930): 117, fig. 4, and another from the
same workshop from Rhodes, Clara Rhodos 8 (1936): 179f., figs. 168, 169; also an example said
to have been excavated at nye on the Black Sea; see E. Akurgal in Antike Kunst 10 (1967):
32 ff., figs. 13, pl. 8. 9: 26, surely from the Achaemenian period and from a local workshop.
26 Dalton, Treasure of the Oxus, p. 10, noted that doubts about the authenticity of the
sheath had been raised, but he says nothing of the gold bowl. I know the vessel only from
photographs but see no compelling reasons to doubt its ancient date. Herzfeld, Khattische
und khaldische Bronzen, p. 154, considered the bowl to be Urartian.
27 Barnett, Median Art, pp. 80 ff., 86 ff.
when properly understood it yields an Achaemenian reference. Further-

more, the addorsed heads are equally relevant for attribution, although they
have nothing to do with Urartian art. The Urartian parallels cited by Bar-
nett, addorsed calves heads employed as clamps for mirror handles, one of
which is in the British Museum, are either Achaemenian or western Ana-
tolian.28 Nor, as Barnett also claims, is the silver dipper excavated at Tell
Fara, which is embellished with addorsed calf heads, an Urartian artifact.
It is more certainly Achaemenian, and the use of calf heads, as well as duck
heads, is a classic Achaemenian motif,29 as is, it need hardly be noted, the
motif of addorsed animals (i.e., on column capitals). All the features and
style of the Chertomlyk dagger point to its Achaemenian background, as
Dalton, Otto, Ghirshman, Amandry, Farkas, and Moorey have already con-
cluded,30 and we need dwell no longer on demonstrating the obvious.
When we turn to the Kelermes (Caucasus) and the Litoy-Melgunov (north
of the Black Sea) akinakes sheaths, we enter a new environment and en-
counter a different set of artistic-cultural problems. Both sheaths are re-
markably similar in shape and in decorative scheme, and it is probable
that they were made in the same workshop. Art historical analysis indicates
this workshop cannot have been an Achaemenian one, either culturally or
chronologically. The shaft of both sheaths is decorated with a frieze of typ-
ical Urartian Mischwesen: on the Melgunev sheath all shoot with a bow; on
28 Ibid., pp. 89 f., n. 2, fig. 5. For published challenges to Barnetts Urartian attribution of
the British Museum mirror, see A. Greifenhagen, Schmuck und Gert einer lydischen Md-
chens, Antike Kunst 8 (1965): 17 f.; A. Oliver, A Bronze Mirror from Sardis, in D.G. Mitten et al.,
eds., Studies Presented to G.M.A. Hanfmann (Mainz: 1971), pp. 117f., 120; Moorey, Cemeteries,
p. 57. The animals on the rim of the British Museum mirror exhibit Achaemenian features,
as Akurgal, Urartische und altiranische Kunstzentren, p. 124, notedalbeit dating it to
the Median period. The Berlin example, published by Greifenhagen, pp. 16f., pl. 5, is, as that
author claims, East Greek in style; it has no addorsed heads as clamps.
29 Barnett, Median Art, p. 90; he was supported in the Urartian attribution of the Tell
Fara dipper by R. Amiran, Two Luristan Fibulae and an Urartian Ladle from Old Excava-
tions in Palestine, Iran. Antiqua 6 (1966): 90 f. For Achaemenian examples, see P. Amandry,
Orfvrerie achmnide, Antike Kunst I (1958); 13 ff., 18; idem, Collection Hlne Stathatos,
vol. 3 (Strasbourg, 1963), pp. 262, 264 f., 267, figs. 158, 163168; Oliver, Bronze Mirror from
Sardis, p. 117; E. Stern, Material Culture of the Land of the Bible in the Persian Period (Jerusalem,
1982), p. 74, fig 90; p. 147, figs. 241, 244; p. 152, fig. 254.
30 Dalton, Treasure of the Oxus, pp. 13, 18; H. Otto, Ein achmenidischen Goldwidder,
ZA 48 (1944): 10, n. 2; 13; Amandry, Orfvrerie achmnide, p. 20; Ghirshman, Ancient Iran,
p. 358; Farkas, Horse and Rider in Achaemenid Art, p. 67; Moorey, Cemeteries, p. 57. Culican,
Medes and Persians, pp. 137, 139, considers the dagger to be Median; Akurgal, Urartische und
altiranische Kunstzentren, p. 124, sees it as a product of medisch-achmenidischen Stiles.
Note that its sheath is decorated in Greek style; M.I. Artamonov, The Splendor of Scythian Art
(New York, 1969), p. 55, pls. 183, 184.
the Kelermes sheath the shooting creatures alternate with passive striding
ones. Also of Urartian style is the pair of winged human figures flanking
a tree at the top of the sheath. On the side projection, however, there is
a reclining stag bordered by stylized eagles beaks, both features of classic
Scythian style, and the chape is decorated with two confronted Scythian
style leonine creatures. The confusion concerning the cultural interpreta-
tion of the sheaths is increased by the material associated with them: at Kel-
ermes, a Scythian axe and at Litoy, Urartian furniture fittings were reported.
Further associated with the Kelermes finds were a gold petaled bowl and
another with repouss animals, both of which seem to be Achaemenian
works.31
All scholars concerned with the sheaths agree on the above description
and analysis; they differ in their interpretation and perception of the artistic
mixtureactually, juxtaposition is the better termand the nature of the
cultural-ethnic workshop that produced them. In 1937 Herzfeld dated the
sheaths to the first half of the sixth century bc, sicher vorachaemenidisch,
dabei so urartaeisch, but in 1941 he placed them in the Median period,
along with the Oxus sheath, and he did not refer to Urartu.32 Barnett devel-
oped and expanded this casual Median period attribution with force, and
again by fiat, in his cited 1962 article. He rightly recorded the Urartian and
Scythian elements but only to announce that the connection of these two
styles plus the akinakes shape signifies we are in the presence of Median
art.33 The Scythian relationship had already been introduced in a casual
manner when, in connection with the Oxus sheath, he referred to the chape
represented on the Audience Hall akinakes.
A close reading of Barnetts conclusions suggests that in fact he began
his studies with the idea that the Kelermes and Litoy sheaths were Median
and then isolated the co-existence of Urartian and Scythian motifs, thereby
producing a formula to demonstrate what already had been determined.
But Scythian and Urartian and Median are discrete cultural entities, two of
which are known, the other unknown, and simple arithmetic cannot yield
an archaeological reality, Median. The unknown has not been revealed.
31 See my Excavated and Unexcavated Achaemenian Art, pp. 25f.

32 Herzfeld in AMI 8 (1937): 135, n. 1.
33 Barnett, Median Art, pp. 82 ff., 90 f.; in Persepolis, on p. 76, Barnett produced a more
ambitious formula or equation for Median art: Urartian plus Assyrian plus Scythian plus
Phoenician elements. What happened to the Assyrian and Phoenician elements in the five-
year interval was not explained.
It has already been noted that the akinakes sheath cannot be considered
to be a Median artifact, an object characteristic of Medes in any period, and
therefore it may not be brought into the discussion as a factor in a Median
equation. Nor indeed may the chape on the Audience Hall relief, which
displays a curled animal in pure Scythian style, for this particular decoration
is an Achaemenian period artifact, as all scholars agree.34 In any event,
the chapes on the sheaths under review here are composed of a different
Scythian form. We have thus to concern ourselves with the iconography and
style of the decoration on the sheaths, and others who have studied them
have arrived at conclusions other than that suggested by Barnett. Aside from
Herzfelds 1937 note, the only other scholar known to me who concluded
that the sheaths were Urartian works, incorporating Scythian elements, was
H. Kantor.35 Others have taken the contrary position, that they are Scythian-
made artifacts incorporating Urartian elements, some singling out the non-
Urartian form of the sheath as crucial to the argument.36
The problem to be investigated here is truly complex, as it involves under-
standing why the motifs of two distinct cultures appear juxtaposed on one
artifact, and some detailed discussion is inevitable. First of all, I do not think
that geography, the location of the find-spots, should play a role in analy-
sis; the final resting place of an object ought not necessarily determine or
34 Idem, Median Art, p. 79. For discussions of these Achaemenian chapes, see the com-
ments of the following; B. Goldman, Achaemenian Chapes, Ars Orientalis 2 (1957); 43

54; R. Stucky, Achmenidische Ortbnder, Archologischer Anzeiger (AA) II (1976): 1323;
P. Bernard, A propos des bouterolles de forreaux achmnides, Revue archeologique 2 (1976):
227246; S. Sorokin, The Curled-up Animal from Ziwiye (in Russian), Soobscheniya Gosu-
darstvennaya Akademia, Hermitazha 34 (1972): 7578; Litvinskiy and Pichikyan, in Acta Anti-
qua 27 (1983): 49 ff., 54. Note that a circular bone plaque excavated at Sardis (BASOR 192 [1966]:
14, fig. 9; BASOR 211 [1973]: 33, fig. 11; Biblical Archaeologist 29 [1966]: 45, fig. 5) seems to be a
solid pommel, not a chape as claimed.
35 H. Kantor, A Fragment of a Gold Applique from Ziwiye , JNES 19 (1960): 13; in n. 49
she cites Herzfelds 1937 reference. G. Azarpay, Urartian Art and Artifacts (San Francisco,
1968), pp, 70 f., 73, believes on the one hand that Urartian models or perhaps even Urartian
craftmanship may have played a role; but then, because the motifs are of mixed style, and
not a normal pattern of Urartian art, they are not Urartian works. But, of course, there is no
mixed style, there are two separate styles. B. Goldman, The Animal Style of Ziwiye, IPEK 24
(19741977): 55, 61, says that the Litoy sheath may be an Urartian import, but on p. 67, he
supports Barnetts views on Median art; a contradiction seems evident here.
36 Barnett, Median Art, p. 91, n. 2; also Ghirshman, Ancient Iran, pp. 28, 303f., 327; Akurgal,
Urartische und altiranische Kunstzentren, pp. 123 f.; Stucky, Achmenidische Ortbnder,
p. 20; Moorey, Cemeteries, pp. 56 f.; B.B. Piotrovskii, Urartu (London, 1967), pp. 21, 97ff. Dalton,
Treasure of the Oxus, pp. xxxv f., considered the frieze to be Assyrian; see also J. Boardman,
The Greeks Overseas (London, 1980), p. 258.
influence a discussion of the cultural background of the artist(s).37 Further,

the position taken here is that art historical analysis per se cannot resolve
the problem of cultural manufacture; we are aware of what is Scythian
and what is Urartian; we are unaware of why or in what manner they
came to be juxtaposed. Thus we are limited to asking certain appropriate
questions, the purpose of which is to define the goal and the problems,
but questions that I believe are not readily answered or resolved except
by inclination. These questions would be asked even if we did not have
Barnetts Median formula before us. When were the sheaths made? Who
commissioned the manufacture of the sheaths, an Urartian, a Scythian, or a
third party? Who staffed the workshop in which the order was fulfilled; was
the artist an Urartian who was trained and skilled in Scythian art, a Scythian
equally trained in Urartian art, or did an Urartian and a Scythian artist work
side by side, one decorating the shaft, the other the added chape and side
projection?
The first question may be answered with little difficulty, for here consen-
sus is more readily achieved than with the others. From what we know of the
chronology of Urartian art, the sheaths were made before the early sixth cen-
tury bc, unless Urartian artists continued to function for new masters after
the collapse of their kingdom, concerning which we know nothing. From the
Scythian side, anytime within the late seventh or sixth centuries bc would
be acceptable. Thus, a time between ca. 650 and 550bc seems to be viable,
and in fact it is a date accepted by all scholars concerned with the sheaths
chronology.
Concerning the second question, there are at least the following possible
resolutions. An Urartian king (or two kings) commissioned the sheaths from
a Scythian (a captive or ally) as an exotic possession, perhaps or perhaps
not assigning one of his best artists to assist. The prominent Urartian motif
along the shaft would in this scheme have been an identity marker on a
non-Urartian object, bearing the message when displayed I am an Urartian
(even though I bear a foreign weapon). Or, equally possible, a Scythian
king could have commissioned the sheaths, presumably a typical Scythian
weapon at this time, perhaps or perhaps not employing an Urartian artist.
Here, however, one must explain why the Urartian scene was so prominently
displayed. In a third scenario, we could conceive of an Urartian or a Scythian
37 Thus Barnett, Median Art, p. 93, explains the presence of the sheaths in Scythian
tombs as booty or gifts from the Medes acquired when the Scythians were in Iran. This
conclusion follows from the view that the sheaths are Median, but if not Median, then
another explanation must be sought for their distribution.
king commissioning the sheaths as gifts from one king to the other, gifts
reflecting the art (message) of both the donor and the recipient.
Finally, a non-Scythian, non-Urartian third-party may have commis-
sioned the sheaths. On the one hand, a king of Land X, itself having or not
having an artistic tradition of its own, may have summoned to his court a
Scythian and an Urartian, or a Scythian/ Urartian trained in both schools,
to create bicultural artifacts for some particular reason (an exotic curio?).
On the other hand, we could conceive the specific thrust of Barnetts con-
cept; a king from a land without artistic traditions may have summoned
the appropriate artists because he intellectually and consciously chose as
the artistic expression of his land (state) the juxtaposition of Scythian and
U rartian motifs, anticipating that the contemporary (not to mention the
future) world would recognize it as such. With the acceptance of either one
of these schemes, another question ineluctably surfaces: which of the many
pre-Achaemenian states in Iran or in the northern regions was the third
party?38
Concerning the workshop staffing, a neutral position recognizes that
archaeological and ethnographical evidence indicates that an artist trained
in one cultural tradition could reproduce the art of another and, equally,
that two artists could work on the same object.39
I have no decisive suggestions to offer concerning the intriguing problem
of sorting out the mental and physical activities involved in the creation
of the sheathsalthough I am inclined to believe that two artists partici-
pated. Further, I find the third-party thesis, in particular that articulated by
38 Barnett (The Treasure of Ziwiye, Iraq 18 [1956]: 114) revealed the characteristics of
Mannacan art as crude, if lively, but barbarous and provincial in the extreme, and he even
presented the archaeological community with actual examples, all picked at random from
the vast Iranian scrapbasket.
39 For an object made in two parts and decorated by two separate artists, see my article
A Bronze Vase from Iran and Its Greek Connections, MMJ 5 (1972): 2550, figs. III. To this,
compare a vessel very similar in form and style, also made in two parts but here apparently
decorated by one artist described by P.R.S. Moorey in G. Markoe, ed., Ancient Bronzes Ceram-
ics and Seals (Los Angeles, 1981), no. 415. Although in the first instance both artists shared the
same artistic tradition, the concept of two artists working on a single object is established
empirically. For ethnographic evidence of this same phenomenon, see M.A. Hardin, The
Cognitive Basis of Productivity in a Decorative Art Style , in C. Kramer, ed., Ethnoarchae-
ology (New York, 1979), pp. 92 f. For artists of one ethnic, cultural background manufacturing
works of another, see the Tomb of Petosiris in Egypt, where Egyptian craftsmen are depicted
making Achaemenian style objects; see my article Excavated and Unexcavated Ancient Near
Eastern Art, in T.C. Young, Jr. and L.D. Levine, eds., Mountains and Lowlands (Malibu, 1977),
p. 194, n. 100. Note that M. Rostovtzeff, Ionians and Greeks in South Russia (Oxford, 1922), p. 50,
believed that one artist fashioned the separate parts of the Kelermes sheath (a Scythian?).
Barnett, to be difficult to accept; it floats without any apparatus of archae-

ological control and is little more than uncharted speculation. But even if
some scholars persist in accepting it, I would argue that it is not viable to
claim that the third party was a Mede to the exclusion of another ethnic
personality. And, to confront one more issue, while the majority of schol-
ars affirm the view that the sheaths were commissioned by Scythians, the
possibility that Urartians were involved cannot be excluded matter-of-factly.
In the final analysis, no archaeologist knows with certainty the answer to
the questions raised; the ancient message, written boldly on the sheaths,
remains at present unread.
We turn now to the so-called Ziwiye treasure. As I have written elsewhere

about the vertiginous problems associated with this treasure, about its
provenience, chronology, perceived artistic backgrounds, and characteris-
tics, I confine myself here to the issue of its alleged role in revealing Median
art. Various scholars have recognized in the Ziwiye treasure the art of the
Mannaeans, Assyrians, Urartians, Cimmerians, Scythians, Babylonians, and
the Greeks,40 but it was M. Falkner, in 1952 shortly after the first publications
on the treasure appeared, who was the first to introduce the Medes. Con-
cerning herself with the background of the artist who created the gold gor-
get, she suggested that Man konnte an einen medischen Knstler denken,
and she cited in support of her judgment the DSf inscription and certain
motifs found in Achaemenian art.41 Ten years later, Barnett, applying his
Scythian-Urartian formula, argued that those pieces in the treasure that con-
tained a combination of these styles, for example, the gold stag and goat
plaque, are Median works. So secure was his conclusion, he found it possi-
ble to claim as an archaeological reality that the treasure derived from the
tomb of a Median chieftain.42
Others accepted this view, more or less. Porada at first considered Bar-
netts Median art/tomb interpretation to be very convincing, but later she
took a more cautious position, deciding it was preferable to reserve
judgement on the label to be given to the Ziwiye treasure. 43 Ghirshman
40 See my Ziwiye and Ziwiye: The Forgery of a Provenience, Journal of Field Archaeology
4 (1977): 197219, esp. pp. 205 ff., 207 f.; see also K.R. Maxwell-Hyslop, Western Asiatic Jewellery
(London, 1971), pp. 206 f.; T. Sulimirski, The Background of the Ziwiye Find . Bulletin of the
Institute of Archaeology 15 (1978): 731; R. Ghirshman, Tombe princire de Ziwiy (Paris, 1979);
and B. Hrouda, Der Schatzfund von Ziwiyah . Iran. Antiqua 18 (1983): 97108.
41 See M. Falkner in Der Schatz von Ziwiye, AfO 16 (1952): 132.
42 Barnett, Median Art, pp. 84 f., 91 f., 94 pl. 4ac.
43 Porada in 7000 Years of Iranian Art (Washington, D.C., 19641965), p. 26; idem, Art of
also shifted his position on the nature of the treasure but more dramati-
cally and in a different direction than Porada. At first he confessed not to
know what the treasure represented, then that it was a Scythian treasure
deriving from a Scythian princes tomb, and still later that it came from a
tomb of a Scythe-Median prince, reflecting here his view that there was
a cultural-political symbiosis of the two peoples (cf. his conclusions about
Luristan, supra). Concomitant with the latter conclusion, he maintained,
with no documentation and without calling attention to a specific object,
that Median art was formulated from many different strainsUrartian,
Assyrian, local Mannaean, perhaps even Ionian . Twenty pages later in
the same publication, after noting that the Ziwiye material has much to
teach us about Median art, he added the oxymoronic non-sequitur it seems
impossible to differentiate the [material culture] of the Medes from the
Scythians and Cimmerians.44 In her book on ancient Near Eastern jewelry,
Maxwell-Hyslop also accepted the conclusion that Median elements are to
be recognized in the Ziwiye corpus. As evidence to support this claim, she
singled out two gold roundels of (to my eyes) Achaemenian style which, on
the basis of comparison with two forgeries (infra), enabled her to conclude
that they could be products of a Median goldsmith.45
We need not dwell any longer on the Ziwiye-Median equation, for the
same arguments, to be sure with more permutations, brought forth in the
discussion of the Kelermes-Litoy sheaths, obtain here, a resort to specula-
tion and singling out of motifs and objects without reference to verifiable
data. And whereas previously a modest equation of which Western Asi-
atic motifs or elements constitute Median art was presented, in the case of
Ziwiye the grouping on the left side of the equation has been unabashedly
modified and enlarged.
About a half-dozen other stray objects and types have from time to time
been labeled Median artifacts. G. Pudelko, writing in the 1930s and ignor-
ing what was even then known about Urartian art, cited a typical Urartian
Ancient Iran, p. 134. For the record, Culican, Medes and Persians, pp. 46f., repeats Barnetts
views; and H.A. Potratz, Die Skythen und Vorderasien, Orientalia, n.s. 28 (1959): 67f., had
earlier connected the Ziwiye treasure to the Medes.
44 For the bibliography on Ghirshmans views, see again my article Ziwiye, pp. 205f.; see
esp. Ghirshman, Ancient Iran, pp. 100, 104, 124.

45 Maxwell-Hyslop, Western Asiatic Jewellery, pp, 210 f. For one of the forgeries, a gold vessel
in the Cincinnati Art Museum, see infra; for the other, see my article Bronze Vase from Iran,
p. 184, no. 156, a gold plaque in the Metropolitan Museum.
winged bull cauldron attachment as eine vielleicht medischer Stierkopf.46

Ghirshman believed that a silver animal-headed vessel from Kaplantu
near Ziwiye was to be attributed sous rserve to the Medians,47 presum-
ably solely because of its apparent date and the geographic location of the
alleged find-spot. E. Akurgal believed he could recognize as coming from
the Median period, and reflecting at least halb medisch elements, the
very same bronze mirror in the British Museum that Barnett had previously
attributed to the Urartians.48 A class of axes with a full bodied lion in relief,
either on the blade or at the socket, one of which is claimed to have been
found at Hamadan, has also been furnished a Median label by R. Dussaud
and Barnett, although, as P. Calmeyer has demonstrated, they are second-
millennium bc products.49
Finally, in this category is a group of six fibulae in the Feroughi collection
in Tehran that warrants some attention. These fibulae have an arc terminat-
ing in a Pazuzu head and are probably of seventh-century bc date. At least
two of these examples have on the opposite arc a clothed female with hands
clasped before the chest and two ram/mouflon heads at the apex on which
the Pazuzu and female stand, and three have a bird opposite the Pazuzu.50
When Ghirshman first published the Feroughi Pazuzu fibulae (in 1964), he
arbitrarily claimed they derived from Luristan, but later (1970) he attributed
them to either Luristan or Kirmanshah, and he suggested that they were
imports from Assyria, Calmeyer originally (1969) supported the Luristan
provenience but later categorically transferred the alleged find-spot of all six
46 G. Pudelko, Altpersische Armbnder, AfO 9 (19331934): 87. Pudelkos Urartian bull
attachment is discussed by G.M.A. Hanfmann, Four Urartian Bulls Heads, Anatolian Studies
6 (1956): 205213, esp. pp. 207 f., no. 2, Louvre AO 17.207.
47 R. Ghirshman, Notes iraniennes XI: Le Rhyton en Iran, Artibus Asiae 25 (1962): 75, 77,
fig. 19.
48 See supra, and n. 28. In The Art of Greece (New York, 1968), p. 217, Akurgal refers to the
object as Median work.

49 R. Dussaud in A Survey of Persian Art (Tokyo, 19641965), vol. I, 266, pl. 52 C. Barnett, Art
of Bactria, pp. 46 f.; P. Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin,
1969), pp. 39 ff.; see also my Excavated and Unexcavated Achaemenian Art, p. 34.
50 R. Ghirshman, La Fibule en Iran, Iran. Antiqua 4 (1964): 97, nos. 1315, pl, 25; idem,
Le Pazuzu et les Fibules de Luristan, Mlanges de lUniversit Saint Joseph 46 (1970): 123ff.,
pl. 2; Calmeycr, Zu einigen vernachlssigten Aspekten medischer Kunst, pp. 114f., D, E, F.
Ghirshmans nos. 13, 14 lack the pin, but cf. Calmeyers E, with the spring on the Pazuzu head.
Calmeyers F consists of only one arc and is odd: there seems to be only one ram/mouflon
head at the apex; a birds head catch is on top of the Pazuzu head (we would expect it on
the opposite arc); and there seems to be no apex. If this object is a fibula, presumably there
was a female at the opposite arc. Nos. 15, D, and F have the ram/mouflon apex heads, nos. 13,
14, and E have the bird.
fibulae to the Kirmanshah area in western Iran. He then proclaimed them to

be Median artifacts, an attribution recently accepted by J. Curtis.51 Calmeyer
argued (if I follow him fully) that the cast-loop to hold chains and seals exist-
ing on some of the Feroughi fibulae, the fact that some have what may be a
stamp seal at the apex, plus the presence of the Pazuzu heads, indicate a
peculiar quality of Iranian fibulae. He further argued that no examples of
this fibula type occur outside of his fictive Kirmanshah area.
Now, fibulae in other areas of the Near East also had apotropaic value;
fibulae with chains and seals occur at Nimrud in Mesopotamia, as noted by
Calmeyer, and at Ras Shamra in Syria, not noted by Calmeyer;52 and fibulae
of the Pazuzu type do occur outside of Iran. As I noted elsewhere in another
context when publishing a related fibula in the Schimmel collection,53 the
type is Assyrian. Moreover, an example with a Pazuzu head on one arc, a bird
on the other, and a stamp at the apex has been excavated, not in Luristan or
Kirmanshah or elsewhere in Iran, but at Megiddo, in Israel.54 Furthermore,
and not insignificant to the argument, the archaeological provenience(s) of
the Feroughi fibulae remains unknown, given their existence in a private col-
lection and their purchase from plunderers. Whether they reached Tehran
as a consequence of recent plunder in Iran, or recent plunder and modern
trade from Iraq or even Israel, can never be revealed by investigation.
At least three of the Feroughi fibulae and the Schimmel example can eas-
ily be defended as Assyrian forms. The examples with ram/mouflon heads
joining at the apex are unique among fibulae and the Pazuzu corpus, and
one tends to think of an Iranian rather than a Mesopotamian background
for the animal heads. I find it difficult to decide with some certainty whether
these examples should be considered as artifacts made in Iran, or if we
51 Calmeyer, Datierbare Bronzen, p. 99, figs. 100, 101: note that fig. 100 is not, to my eyes,
Ghirshmans La Fibule en Iran, (1964), no. 14, as is claimed; I do not know what this example
is (a female head and a Pazuzu head on opposite arcs); idem, Zu einigen vernachlssigten
Aspekten medischer Kunst, pp. 114 ff.; Curtis, Nush-i Jan III, pp. 23, 33.
52 See my article Fibulae Represented on Sculpture, JNES 26 (1967): 82ff. For fibulae
associated with chains, see M. Mallowan, Nimrud and Its Remains, vol. 1 (London, 1966),
pp. 114 f., fig. 58; C. Schaeffer in ILN, April 27, 1935, p. 686, fig. 2; for two examples from
Luristan (excavated!), see L. Vanden Berghe, Les Fibules provenant des fouilles au Pusht-i
Kuh, Luristan, Iran. Antiqua 13 (1978): 51, nos. 10, 11, fig. 9, pl. 2.
53 See my article Five Additions to the Norbert Schimmel Collection, Acta Praehistorica
et Archaeologica 7/8 (19761977): 316, no. 3. The Schimmel fibula has a female on each of the
two arcs; they are of the same style as those on the Feroughi fibulae.
54 Amiran, Two Luristan Fibulae and an Urartian Ladle, pp. 88ff., pl. 17, I; Amiran
recognizes the Megiddo fibula to be an import there, but uncritically following Ghirshman,
she claims the import came from Luristan.
have before us merely another form or variety of an Assyrian fibula. If made

in Iran, indeed a possibility albeit not archaeologically proven, they would
be local copies of an Assyrian fibula with local modifications. But if such
be the case, what local culture decided on the modifications? To name the
locality in question, if indeed it is an actual issue, to maintain that these
two fibulae or, with Calmeyer, all six of the Feroughi fibulae were made by
Medes and no other people, is to abuse the evidence, which consists of six
unexcavated objects in Tehran and one excavated example from Megiddo.
If the previous group of stray Near Eastern objects cannot meaningfully
be accepted as Median artifacts, the next group to be considered cannot
intelligibly be accepted as ancient Near Eastern artifacts, let alone Median
ones. I am aware of at least seven objects, all unexcavated, all presented to
the archaeological community by antiquity (?) dealers, that are manifestly
not the products of an ancient Near Eastern artist but notwithstanding have
been baptized by various scholars as examples of Median art. I hold the
view that in each case the forger was in fact attempting to create a work
in the style of Achaemenian art, but he misunderstood and failed to attain
the correct subtle form and style of that art and, in a few adventurous cases,
created forms not known in that art. The forgers knew nothing of Medes,
but they knew the market value of Achaemenian antiquities, and they knew
from past experience that scholars would welcome their merchandise. So
eager has been the desire to isolate Median art from the mass of disparate
material said to come from Iran, so eager the attempt to recognize what
is surely there, that these artifices have been transmogrified by ingenuous
scholarship into pre-Achaemenian art. Rarely has it been considered as
a possible interpretation that these works, shouting their deviation from
Achaemenian style (a fact noted by all commentators), might be modern
creations, post- and not pre-Achaemenian.
The best-known, because the most cited, example of Median art (ein
charakteristischer Vertreter der vorpersischen Kunst) is a gold vessel in
the Cincinnati Art Museum. The vessel has a sharply carinated body and
two janus-headed horned animal handles, and it is, appropriately, claimed
by its original publisher to have been found at Hamadan.55 For the shape
55 See again my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 152;
and idem, Addenda to this article (Malibu, 1979), p. 4, no. 24, for the bibliography of scholars
who published this piece as genuine: now add Akurgal, Urartisch und altiranische Kun-
stzentren, p. 123 (also for the quote in my text); Ghirshman, Tombe princire de Ziwiy, p. 21;
G.G. Beloni, Iranian Art (London, 1969), pl. 40. Cf. P. Calmeyer, Hamadan, in RLA, vol, 4, p. 66
[note that in my article Excavated and Unexcavated Achaemenian Art, in nn. 12, 33, above,
and configuration of the vessel, for the sharp, stylized, and poorlyinaccu-
ratelyexecuted decoration of winged lions in relief, there is no parallel
in Achaemenian or ancient Near Eastern art; and the feet of the animal
handles rest directly on the vessel, contrary to proper Achaemenian prac-
tice. Another gold vessel of similar form and relief decoration, this one
in a private collection, has lion handles, and its shape and execution are
equally unparalleled in style and execution in Achaemenian and ancient
Near Eastern art. Here there is an additional motif of two lions sharing
one en face head, a motif found on still another forgery.56 To defend these
two manifestly modern misunderstandings, to ignore the loud and clear
message of an unexcavated object supplied with a forged provenience, is
to mock proper archaeological and art historical analysis. A third vessel of
gold, an animal-headed cup now in Tehran, is an object that I also believe
is not ancient; Ghirshman considered it to be ancient and sous rserve
Median.57
A gold strip in the British Museum depicting a winged lion in relief is also
to my mind not the work of an ancient artist, although claimed as Median by
Barnett; nor is a gold handle with addorsed lapis lazuli lions at the top, also in
the British Museum, to be considered the work of an ancient artist.58 A silver
bowl in the Ashmolean Museum decorated in the interior with two crossed
lions was originally published as Median and from Hamadan; subsequently
one of its publishers has recognized it as a modern work.59
Finally, in the Louvre is a fragment of a white stone head executed in
barely recognizable Achaemenian style. A. Parrot and P. Amiet have identi-
fied it as most probably a Median work because its character is just not quite
Achaemenian in either style or execution: the hair, beard curls, mustache
I incorrectly stated that Calmeyer in this publication claimed that the gold Achaemenian
objects in the Metropolitan Museum derived from Hamadan. In fact, on p. 65, left, he raised
doubts about this attribution. I here correct my error and apologize; I apparently misread my
notes].
56 See again my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 153;
cf. no. 156 (see n. 45 above); see again my Addenda, p. 4, no. 25. Note that the roundel with
the head shared by two lions discussed by Maxwell-Hyslop, Western Asiatic Jewellery, p. 210,
may be genuine; see another probably genuine example of this motif, this time on a pre-
Achaemenian, Luristan disc pin, Ghirshman, Ancient Iran, fig. 384.
57 Ghirshman, Notes iraniennes XI, pp. 75, 77, fig. 18; see again my article in Young and
Levine, eds., Mountains and Lowlands, p. 185, no. 166; and my Addenda, p. 4, no. 29.
58 For the gold strip, see Barnett, Median Art, p. 78, pl. 1a; and my article Ziwiye, p. 211.
The gold and lapis handle (135908) has been published as a postcard dated 1975: Median or
Early Achaemenian.
59 See my article in Young and Levine, eds., Mountains and Lowlands, p. 184, no. 154.
positioning, eyes, ears, nosein short, everything about the head is wrong.60
Furthermore, there is a debate between the two scholars involved in the
heads presentation concerning whether, inasmuch as the provenience is
accepted as Hamadan (although purchased in Switzerland!), it is a portrait
of Cyaxares or Astyages, a thought that surely did not enter the mind of the
forger. The Louvre head joins the other forgeries of Achaemenian objects
discussed above in the repertory of new Median art.
We turn now to the final group of objects encountered in the Median
odyssey, works that have actually been excavated in Iran itself. At the pres-
ent time, perhaps the least controversial element in the alleged Median
corpus is the group of rock-cut tombs: Dukkan-i Daud, Farhad-u-Shirin,
Sakavand, Fakhrika in Iran, and Kizkapan in Iraq. Herzfeld gave them promi-
nence as Median tombs because of their geographical positions and because
of the Median dress on some of the figures in relief on the tombs at
Dukkan-i Daud and Kizkapan, to which he compared the gold plaques from
the Oxus treasure (supra); he dated Sakavand to the early Achaemenian
period, to the year 521 bc.61 A few scholars, including I.M. Diakanov, H.H. von
der Osten, and Ghirshman, also believed the tombs to be Median, but recent
scholarship has rejected this early chronology. Research based on more
refined dating of tool marks and stylistic analysis of the columns before
some of the tomb openings indicates that they are Achaemenian or later.
We still seek the tombs of the Medes.62
When Stronach originally published a bronze Pazuzu head excavated
at Nush-i Jan, he considered it not at all unlikely to have been looted
from Assyria.63 Now, whether it was looted from Assyria or thence derived
by more peaceful means is not of course known, but its Mesopotamian/
Assyrian origin is almost certain.64 A Mesopotamian attribution and origin
60 A. Parrot, Acquisitions et indits du Muse du Louvre, Syria 44 (1967): 247ff.; Amiet
is quoted in the text. I regret not including this piece in my 1977 paper on forgeries (in
Mountains and Lowlands) but correct the omission here; one need only read Root, King and
Kingship, pp. 114ff., to be convinced of its modern birth.
61 Herzfeld, Iran in the Ancient Near East, pp. 200ff.
62 I.M. Diakonov apud R. Ghirshman in review in Bibliotheca Orientalis 15 (1958): 261; von
der Osten, Die Welt der Perser, pp. 56 f.; Ghirshman, Ancient Iran, pp. 87ff. For the lower
chronology see H. von Gall, Zu den medischen Felsgrbern in nordwestiran und iraqi
Kurdistan, AA 1 (1966): 19 ff.; Porada, Art of Ancient Iran, pp. 138f.; D. Stronach, Pasargadae
(London, 1978), pp. 283 f.; idem in Achaemenid Village I at Susa and the Persian Migration to
Fars, Iraq 36 (1974): 246 f.; de Francovitch, Achaemenid Architecture, p. 257.
63 D. Stronach, Tepe Nush-i Jan: A Mound in Media, BMMA 27 (1968): 185, fig. 14.
64 For discussion and bibliography of Pazuzu heads see M. Noveck in Oscar White Musca-
rella, ed., Ladders to Heaven (Toronto, 1981), pp. 139 f., 318f.
were nevertheless rejected by Calmeyer, who, because he had demonstrated

to his satisfaction that the six Feroughi Pazuzu fibulae were made in west-
ern Iran by Medes (supra), added the Nush-i Jan Pazuzu head to the same
Median repertory.65 In the recent publication of the small finds from Nush-i
Jan, Curtis expressed agreement with Calmeyer, and he also employed the
Feroughi fibulae as supporting evidence.66 An examination of the form and
style of the head reveals to me, as well as to D. Stronach, no feature or charac-
teristic that distinguishes it from other Pazuzus excavated in Mesopotamia.
Consequently, to label a typical Mesopotamian artifact as Median begs
many questions and implicitly denies the possibility of recognizing the dif-
ference between Mesopotamian and Median art: if the Nush-i Jan Pazuzu
head is Median, which Pazuzu heads are Mesopotamian? Moreover, even
were we to assume, or guess, that the head was a locally made copy of an
Assyrian original, a conclusion Curtis is willing to accept, the head is a faith-
ful copy. Accepted as such, the head could be understood as an artifact
apparently used by Medes at Nush-i Jan but not as an example of Median
art, a crucial distinction, indeed.
Influential in the attempt to interpret the Nush-i Jan Pazuzu head as
Median is the combination of geographical and chronological data, the
focus being an object of a certain date excavated at a Median site. The
methodology is of course correct, although in this instance, it is argued,
misapplied. A broad geographical-chronological approach was also under-
taken with respect to another object excavated at a Median site. In 1914
French archaeologists excavated briefly at Hamadan, the capital city of the
Medes. Among the finds were fragments of a bronze vessel with a curved
spout terminating in a lions head and with a plaque attachment at the rear
of the vessel depicting a winged male figure; it was probably made some-
time between the late ninth and the seventh centuries bc.67 Several other
examples of this vessel form, with either the same curved spout or a vertical-
horizontal one and with a figured plaque attachment, are known courtesy
of plunderers, and the Ramadan vessel remains the sole excavated exam-
ple.68 In addition to this form, a number of other metal spouted vessels
65 Calmeyer, Zu einigen vernachlasslgten Aspekten medischer Kunst, pp. 114ff.

66 Curtis, Nush-i Jan III, pp. 23, 33 f.
67 Calmeyer, Zu einigen vernachllssigten Aspekten medischer Kunst, pp. 113f., figs. 35;
see again my Excavated and Unexcavated Achaemenian Art, p. 31, figs. 7, 8.

68 For other spouted vessels with figured rear plaques, see Calmeyer, Datierbare Bronzen,
p. 100, fig. 104, Cc, 102, Au; J. Waldbaum, Luristan Bronzes, Record of the Art Museum Princeton
University 32 (1978): 9, figs, 2, 3; P. Amiet, Les Antiquits du Luristan (Paris, 1976), no. 90. See
with vertical-horizontonal spout and without the figured plaque have been
excavated in Iran at Marlik Tepe, Sialk B, and Tepe Guran; and many stray,
unexcavated examples, all attributed to Iran, exist in collections.69 There
is no doubt that these metal spouted vessels, both figured and plain, are
Iranian, that is, they were made somewhere in Iran.
Now, whereas in 1969 Calmeyer stressed a Luristan provenience for the
majority of the stray vessels, in 1974 he played down this alleged provenience,
stressed the importance of the Ramadan find, and concluded that the sites
where the vessels were found (gesicherten Fundorte) constitute a Gebiet
das im 8. Jahrhundert nur medisch sein kann; therefore he believed that
the vessels are Median.70 Notwithstanding this claim, the places where the
plain metal vessels were excavated encompass a wide area in Iran, and
it is archaeologically not possible to assign them to one manufacturing
or cultural center. As for the Hamadan vessel with its figured plaque, it
stands alone without reference to any fixed datum that would allow it to
be considered as either an import or a locally made work. The style of
the figured plaque is best described as western Iranian, a general term
indicating that there is insufficient data to realize cultural differentiation
among the many first millennium bc artifacts deriving from western Iran.71
Indeed, it is not impossible that the Hamadan vessel is a Median work of the
pre-unification period, but I would argue that the Median label should be
kept in parentheses, so to speak, and that at present the vessel be considered
simply a product of western Iran.
Calmeyer is aware that his identification of the Pazuzu fibulae, the Nush-i
Jan Pazuzu head, and the metal spouted vessels as artifacts made by Medes
does not thereby furnish scholars with a clear-cut, unified Median style;
they are not products of a Staatkunst,72 Inherent in this conception is its
also The World of Persian Pottery: Gluck Collection (Tokyo, 1980), no. 66, which (if genuine)
has some form of plaque at the rear.
69 See Calmeyer, Datierbare Bronzen, pp. 99 ff. for the references; unfortunately, he treats
the unexcavated examples as if they were excavated, assigning a Luristan provenience to

many and accepting other unexcavated examples as deriving from Tepe Giyan and Khurvin;
his alleged distribution map on p. 104, fig. 107 is therefore skewed. A number of plain metal
spouted vessels have been excavated at Hasanlu, remain unpublished.
70 Calmeyer, Zu einigen vernachlssigten Aspekten mediseher Kunst, p. 114; idem, Da-
tierbare Bronzen, pp. 149 f.; Hrouda, Vorderasien, vol. 1, p. 286.

71 P.R.S. Moorey, Ancient Persian Bronzes on the Adam Collection (London, 1974), p. 72,
notes that the figured plaques resemble the figures holding two animals at bay on a non-
Luristan cheekpiece, no. 37A.
72 Calmeyer, Zu einigen vernachlssigten Aspekten medischer Kunst, pp. 116f.
own dissolution. If modern scholars cannot isolate and articulate about

Median style and art, if Median art never achieved a self-evident style or
iconography, manifest either on mundane or sumptuous works, how then
can an archaeologist recognize an object as Median? Logically extended,
Calmeyers conception actually permits any first millennium bc object,
presumably limited to those excavated in or said to come from Iran, to be
identified by personal intuition as one made by Medes: precisely what has
in fact been accomplished over the years.
Scholarly research on Median art is somewhat in the same situation as
that concerned with Hurrian art.73 Archaeologists and art historians, know-
ing it must exist, seek it and find it; the topos is the same in both cases. Both
nevertheless remain elusive, paradoxically outside of our present compre-
hension. If any of the above discussed objects or still others are Median-
made, Median art in some ideal historical reality and independently of
the rejections presented here, they are so without our awareness and exist
unperceived by the modern world.74
73 For a discussion of Hurrian art and the impossibility of recognizing and defining it, see
M.T. Barrelet et al., Mthodologie et critique, I: Problmes concernant les Hurrites (Paris, 1977).
The Urkish lions (about which I plan to write elsewhere) have a bona fide Hurrian inscription,
but stylistically they are not distinguishable from Akkadian or Ur HI art of the second half of
the third millennium bc. As such, the lions epitomize the problem of identifying Hurrian art,
distinct from the art of other Near Eastern cultures.
74 Alas, the negative aspects of this study do not end, for not only are Median art and
artifacts unperceived in our time, so are the Medes themselves: in the sense that (leaving
Persepolis aside) not a single figure depicted in Near Eastern art can be recognized with
security as a Mede, a Mede strictu sensu, not a Zagrosbewohner. I make this strong statement
with ease, and in spite of the fact that a number of scholars think otherwise and have
presumed to recognize Medes represented in art at Ziwiye, on stray objects, on Assyrian
reliefs, on Nimrud ivories, viz., Barnett, Iraq 18 (1956): 116; idem, in SPA 14 (1967); 3,000f.,
figs. 1,0611,063; von der Osten, Die Welt der Perser, p. 57; C.K. Wilkinson, More Details
on Ziwiye, Iraq 22 (1960): 217; Culican, Medes and Persians, p. 148; Mallowan, Nimrud and
Its Remains, p. 250, fig. 215; Ghirshman, Un Mde sur les bas-reliefs de Nimrud, Iraq 36
(1974): 37 f., pls. 3 and 4; Sulimirski, Background of the Ziwiye Find. , p. 26; Root, King
and Kingship, pp. 115, 260; The Metropolitan Museum of Art Guide (1983): 57, no. 36, On some
of these alleged identifications, all achieved by guesswork, see the comments of M. Wfler,
Nicht-Assyrer neuassyryscher Darstellungen (Neukirchen-Vluyn, 1975), pp. 118, n. 601; 210,
n. 1,089; 277, n. 1,408. Even the very appearance of the Medes is denied us!
My aim in writing this paper has been to share the results of an examination of the claims
made over the last eighty years that Median art is an entity capable of being recognized by
modern scholars. My conclusions are entirely negative with respect to these claims, but I
strongly believe that the results are positive in the sense dmolir pour mieux btir. Perhaps
now that the table is cleared, we can start afresh, beginning with the questions of what is
Media, what is Median. Perhaps only then can we confront the other question, why is there
no Median art?
chapter thirty-nine
MUSEUM CONSTRUCTIONS OF THE OXUS TREASURES:

FORGERIES OF PROVENIENCE AND ANCIENT CULTURE*
Ia. The British Museums Oxus Treasure (OT-1)
In 1905 O.M. Dalton, then a Keeper/Curator at the British Museum, pub-

lished The Treasure of the Oxus (reprinted in 1926 and 1964; the latter, edited
and altered by R.D. Barnett, is the edition cited here) on a remarkable col-
lection of gold and silver artifacts acquired by the British Museum in 1897.
Known in scholarly literature for over a century as the Treasure of the Oxus,
or the Oxus Treasure, and dated by scholars either from the 7th century bc,
Median, or the Achaemenian, through the later centuries bc, the collec-
tion is a corpus of about 180 objects. Among these are gold bracelets and
armlets, a silver caprid vessel handle, gold discs, rings, vessels, earrings and
pendants, a gold sword scabbard, two model gold chariots containing three
gold statuettes,1 several silver and three more gold (including a priest and
a horseman) statuettes, gold vessels, two stone cylinder seals, about 50 dec-
orated small, thin flat gold plaques, and many coins.2 Dalton recorded but
did not illustrate each object; this was achieved in a fully illustrated cata-
logue written by E.V. Zeymal (1979) for an exhibition at the Hermitage. The
collection has been hailed by a number of Keepers at the British Museum:
viz. by Barnett as one of the treasures of the British Museum and the British
Nation, and later as one of the treasures both of the British Museum, and
the archaeological world, which cultural evaluation was later reprised by
J. Curtis.3
* This article originally appeared as Museum Constructions of the Oxus Treasures:
Forgeries of Provenience and Ancient Culture, Ancient Civilizations from Scythia to Siberia
9, nos. 34 (2003): 259275.
1 One is the chariot box occupant mentioned both by Cunningham 1881, 151, 154, pl. XII.
no. 8 (as first in the collection of a Louis Cavagnari) and Dalton 1964, vvi, xl. fig. 21 as in
the collection of Lord Lytton; it was donated to the British Museum in 1953. The British
Museum example seems to be Cunningham page 154, pl. XII. no. 8; the gold horseman
statuette donated by Lord Lytton in 1931, Barnett-Dalton 1964, vvi, the plate supplement,
is Cunninghams 154, pl. XIII. no. I.
2 Dalton 1964, 139, pls. IXXI; Curtis 1997, 230233.
3 Dalton 1964, v; Barnett 1968, 34; Curtis 1997, 231.
1026 chapter thirty-nine
What appropriately may be called the official Museum documents of the

discovery is presented by Dalton, a record based on reports that in every
instance were provided by individuals who in turn based their reporting on
second-orthird-hand sources.4 In most cases Dalton gives no references for
the discovery other than the published reports of a military resident, Gen-
eral A. Cunningham in 1881 and 1883; he did not mention other early sources
later reported by Curtis.5 The story told was that the treasure was found on
the Oxus [River] in the year 1877, but the precise site is variously given,
i.e. different sites were named as the find spot (see below); and it was said
to have been found in the sands. The treasure languished somewhere for
three years, for we are told that it was in 1880 that a Captain F.C. Burton6
came across information that some merchants traveling from Bokhara had
been robbed of a quantity of gold near Peshawar, in northwestern Pakistan.
In exemplary fashion Burton tracked down and rescued both the objects
and the kidnapped merchants to whom he was able to return about three-
quarters of what had been robbed (most of which was in sealed bags and
not visible to him). It was from these merchants, who wrote a deposition on
the matter, that Burton learned that they had purchased the objects which
had been found at a site on the Oxus.
Dalton could regret that no qualified observer was present at the find,
but in the very same sentence easily record but fortunately we have sat-
isfactory evidence that the treasure was actually brought out of the Oxus
valley into Kabul, and thence taken to Peshawar, where it was sold.7 Pro-
ceeding to Peshawar (or in a contradictory statement on p. xv to Rawalpindi,
almost 100 miles distant) in (modern) northeastern Pakistan, the Bokhara
merchants sold their rescued Oxus treasure to other merchants. Over some
time the material was purchased in Rawalpindi by Cunningham and A. Wol-
laston Franks; Dalton writes as if these two purchased the whole merchants
lot. Curtis sidesteps the issue of the two distinct market bazaars mentioned
by Dalton and states that most of the treasure ended-up in Rawalpindi.8
Cunningham eventually sold his share of the purchased goods to Franks,
who donated his purchases to the British Museum in 1897.
Barnett summarized the same tale (but, pace Barnett, no one claimed
that Burton was alone when he allegedly confronted the robbers), noting
4 Dalton 1964, xiiixv, and Curtis 1997, 235236.

5 Curtis 1997, 238239.
6 For Burton see Curtis 1997, 248, note 51.
7 Dalton 1964, xiii.
8 Curtis 1997, 238.
forgeries of provenience and ancient culture 1027
the fact that Cunningham had gradually bought his objects over time.9 Dal-
ton and Curtis also noted that Franks purchased Oxus material directly
from local vendors as well as from Cunningham.10 There was no single ven-
dor for the British Museums Oxus Treasure, nor, as Curtis has established,
only one account for the robbed merchant story, or evidence for the actual
existence of Burtons deposition.11 No evidence exists that Dalton ever met
Burton, and we do not know where the former got his information. Cun-
ningham mentions a Major Burton, not with regard to any story concerning
the discovery of the treasure, but solely as someone who owned Oxus
objects.12 Barnett said of Burtons story that it suggests more readily the
atmosphere of romance rather than the sober world of archaeology; and
Pichikyan excitedly and positively described the adventures as bordering
on fiction,13 Indeed.
Dalton also notedbut again avoided the logical archaeological implica-
tions of his statementsthat in the bazaars of Rawalpindi (and Peshawar?)
there had been addition and interpolation to the original Oxus collec-
tion; that dealers in Northwestern India (Pakistan today) frequently receive
antiquities of various periods discovered within and beyond the frontier [ital-
ics mine] and incorporate miscellaneous objects of various origins in
some find 14 His concerns were specifically expressed with the 1500 coins
that arrived with the Treasure, calling attention to the fact that there is no
absolute certainty that they all were found with the treasure, and therefore
they could not function as chronological markers. Put another way, he was
plainly reporting here that the Oxus treasure the Museum purchased as
deriving from one locus had been salted. Dalton was admittinghowever
obliquelythat the treasure curated by him as from a single find was not
from one find, was not what he was claiming it to be. He was also aware that
in 1883 Cunningham wrote about coins be had seen only recently, i.e. years
after those he saw and reported in 1881 and which he continued to recognize
as deriving from the very same locus as those found in 1877.15
Dalton could not allow himself to dissociate the coins from the treasure;
and he did not mention that a numismatist had earlier challenged their
9 Barnett 1968, 3637; also Curtis 1997, 234.

10 Dalton 1964, xvi; Curtis 1997, 242.
11 Curtis 1997. 239240.
12 Cunningham 1881, 183; idem 1883, 65.
13 Barnett in Dalton 1964, Preface; Pichikyan 2002, 213.
14 Dalton 1964, xvxvi.
15 Cunningham 1881, 152, 162184, 258; idem 1883, 64.
inclusion in the treasure.16 Nearly all scholars working with Daltons account
did not perceive the implications of his caution about the multiple origins
(and thereby their obvious absence of unity) of the coins (not to mention the
other objects in the treasure), blithely accepting them as part of the treasure.
Collon expressed initial doubts about the coins belonging to the treasure
but dropped it.17 And more recently, Curtis stoutly defended them as bona
fide components of the treasure; he cited Cunninghams claim/defense on
the matter, which was based on nothing more than the statements of the
collectors themselves, i.e. the parti pris plunderers and the bazaar vendors.18
Curtis understood Daltons abeyancethe origins of the Treasure seem to
be fairly diverse, and followed his lead, requesting that scholars accept the
coins as a component of the treasure. With more import than he realized
he wrote it does seem illogical to accept the Treasure itself as a group
and to throw out the coins. Quite so. But the argument expressed here also
expresses the reverse: it seems illogical to throw out the coins and accept
the treasure as a group, as Dalton and Curtis most surely understood, for
the coins are in precisely the same unprovenienced situation as every other
object in the treasure. Both museum curators therefore saved the coins
hoping to convince scholars thereby of the archaeological integrity of the
complete treasure housed in the British Museum.
In fact all the information assumed to be known about the derivation of
the Oxus Treasure, which has obtained the recognition of archaeologists
and scholars, or general agreement19 derives from Cunningham, Franks,
and others noted by Curtis, and also Daltonthe latter two making archae-
ological determinations about the unity, the Oxus locus, of the Museums
collection from London. But the reality remains to instruct us that no one
writing about the Oxus treasure, and this includes Burton and Cunning-
ham, had any personal experience or knowledge whatsoever concerning
the find spot(s) of the purchased objects about which they reported. Cun-
ningham who gradually bought objects in bazaar stalls situated hundreds
of miles distant from their alleged find spot, and years after the find date
of the recovery reported to and by him, could not have confirmed any-
thing about the origin of the treasures he collected.20 Nor can any modern
scholar confirm merely by citing others who themselves cannot confirm
16 Reported by Curtis 1977, 234235.

17 Collon 1995, 185.
18 Curtis 1997, 233.
19 Dalton 1964, xvi; Curtis 1997, 235236.
20 As Curtis 1997, 235.
anything they report. What remains completely unknown is (1) where and
when the alleged Bokhara merchants acquired their (mostly unidentified)
alleged treasure goods, (2) what role the Peshawar/Rawalpindi bazaars
played in the assemblage of these very goods, and (3) whether the objects
purchased by Cunningham and Franks were in fact those carried by the
Bokhara merchants.21 Porada put it quite neatly and correctly: Daltons story
was a reconstruction22and an inadequate one at that.
A believers anticipated claim that surely the objects must have come
from somewhere nearby is certainly possible, but from an archaeological
perspective is culturally meaningless. I have visited shops in Van, Turkey,
that sell goods from all over Turkey, the United States, Europe, Iran and Iraq;
in Ankara shops I have seen offered for sale material from the Caucasus, Iran,
Uzbekistan and other Central Asian states. The genuine objects sold in the
Peshawar/Rawalpindi Pakistan bazaar could have derived from anywhere in
Pakistan, India, Iran, Tajikistan, Uzbekistan or Afghanistan, yes, including a
site or sites on the Oxus River. Cunningham (who was a learned scholar)
and Franks were not liars; they merely reported what they had heard from
dealers. They were not the last persons to do this. And it is well known that
to this day dealersincluding those at Peshawarsell forgeries and supply
findspot names for the objects they sell to eager curators and collectors
who have to have them. Dalton did what he believed every curator was
supposed to do, defend the material in ones prestigious institution against
all non-evidence or evidence to the contrary; he saw it as his responsibility.
In several publications I have argued against accepting the Oxus Trea-
sure as a bona fide find from one specific site or even one general area,
based on some of the reasons given above.23 This position is beginning to
gain support;24 but, generally, scholars have rejected it. In addition to Cur-
tis, I.R. Pichikyan mounted a vigorous defense of the integrity of an Oxus
locus.25 Pichikyan knew that the objects (all of them) derived from one site.
But more: he also knew their precise locus, which was at Takht-i Sangin, a
temple site he was excavating on the Oxus (Amu Darya) River, in Tadjik-
istan: notwithstanding the reality that none of the Takht-i-Sangin artifacts
relate to those in the British MuseumCurtis can only note that three gold
21 Bleibtreu 1998, 145.

22 Porada 1965, 242, note 45.
23 Muscarella 1980, 26; idem 1987, 113, note 14; idem 2000, 208, note 30.
24 Bleibtreu 1998, 146147. I will not list here all the scholarly references over time to the
Oxus Treasure.
25 Curtis 1997he does not mention my 1980 and 1987 articles; Pichikyan 1992, I, 31ff.,
42ff., 60 ff., 77 ff., 95, contra Muscarella 1980, 1987.

plaques are broadly comparable to the fifty or so examples in the Oxus Trea-
sure.26 P. Bernard, who supported the integrity of the British Museums trea-
sure and the general area from which it supposedly derived, also supported
Cunninghams locus, claiming contra Pichikyan that the treasure derived
from Takht-i Kuwad (spelled different ways), near Takht-i Sangin.27 E. Rehm
supports Pichikyans view; Curtis, citing also Zeymals support for Takht-i
Kuwad, reserves judgement.28
The Oxus Treasure conception was generated by bazaar archaeology
methodology and scholarship. The objects ascribed to it were found (most
certainly not all, see below part II) by a number of individuals and sold by an
unknown number of bazaar merchants to a number of Europeans over sev-
eral years. It is troubling to read Curtis inconsistently first acknowledging
Cunninghams and Daltons reporting the different dates for the find, and
(correctly) concluding: It seems certain, then, that the Treasure was not
all found at the same time, and neither was it found in exactly the same
place, and then two pages later concluding the very opposite: that the evi-
dence makes it quite clear that the Treasure [all the objects in the British
Museum] was found on the banks of the Oxus.29 Something is not quite right
here: hearsay about a find spot (time and place, when and where) equals a
known find spotthis is an archaeological oxymoron (an oxusmoron). Pace
Curtis, the British Museums treasure cannot be accorded archaeological,
cultural or historical value for it is nothing more than a museum-culture
artifact, from time to time polished and glamorized with museum-speak
rhetoric and defenses. Only with irony can it be proclaimed that Cunning-
ham and Franks saved the Oxus Treasure for posterity.30 Historical reality
acknowledges only two facts: Cunningham and Franks purchased over time
individual related and unrelated objects, and, in collaboration with museum
curators, created an Oxus Treasure.
There may be an Oxus Treasure but there is no actual Oxus Treasure,
no hoard from one site wonderfully preserved by many extraordinary indi-
viduals. All discussion must thus commence from the understanding that
each museum provenanced treasure object exists as an unprovenienced
unit, isolated even from any other unprovenienced unit, those alleged (with-
out evidence) to be associated with it. Stripped of the rhetoric, the Oxus
26 Curtis 1997, 237.

27 Bernard 1994, 101 ff., 103, 106.
28 Rehm 2000, 629; Curtis 1997, 236237.
29 Curtis 1997, 234, 236.
30 Curtis 1997, 131.
Treasure, is revealed to be but another member of a very large museum-

collector community populated by objects said to come from various sites;
it is but another forgery of a provenience. Witness the forged site prove-
niences of many hundreds of diverse unexcavated ancient and modern arti-
facts that are assigned by bazaar merchants, curators, and innocent schol-
ars to Ziwiye, Marlik, Luristan, Kalmakarra Cave, Urartu, Dorak, Hamadan,
Afghanistan and so forth.31
Ib. The Miho Museums Oxus Treasure 2 (OT-2)
In a long article written in 1997, Pichikyan returned to Oxus issues again, but
with a renewed and heightened fervor, for his defense this time was in the
context of another significant Oxus manifesto. He first reaffirmed his belief
concerning the Oxus locus of the British Museums treasure, its authorized
history, and its obvious relationship to his site at Takht-i Sangin. Then, with
undisguised passion he shared with his colleagues an amazing discovery he
made. This was seeing with his own eyes in the Miho museum in Japan a
component, the second part, of the British Museums Oxus Treasure, which
he proclaims is the lost and now recovered Oxus treasure (OT-2). It is the
greater part, of that treasure, the lesser being in the British Museum, which,
because of priority of recovery, is now to be known as OT-1.32 The Miho
museums find [purchase] stems precisely from the Oxus treasure and not
any other hoard [italics mine] which, along with OT-1, is Pichikyans temple
at Takht-i Sangin.33 Using the examples of Daltons Oxus treasure story (and
Indiana Jones movies) as exemplars, Pichikyan purports to report on the
archaeological discovery of the OT-2 treasure. An individual had acquired
the major part of the treasure that was found in 1877. He buried his share
of the material soon thereafter, but died before revealing the burial spot to
his children. From that time, his sons and grandsons (great-grandsons are
not mentioned) continuously searched for it and after 120 years of patient
digging the grandsons found the buried hoard! Thereupon, they somehow
got the mass of material to Peshawar thence to Englandthe old OT-1
trailand finally to its host country, Japan.
31 See for example a gift Wollostan Franks donated to the British Museum, an axe he said
was found at Hamadan in 1880, a provenience Barnett accepted and passed on as gospel,
Barnett 1968, 4647, note I, pl. VII, 5; for a summary of some of the provenience issues raised
here see Muscarella 2000, 3132, 5361, 76, 8183, 141, 146147.
32 Pichikyan 1997, 306307; see his note 2.
33 Pichikyan 1997, 308310.
What the archaeologist Pichikyan insists, straight faced, is that schol-

ars recognize the reality of the archaeological find he reports, the lost
treasurethe second part of the Oxus Treasure.34 Archaeologists must be-
lieve that these long-sought artifacts had been recovered together with OT-
1 from one site on the Oxus River in the 19th century, that all the Miho
OT-2 artifacts were recently recovered, carefully kept together, transported
great distances (of course legally) from their country of origin, and then
brought to England to be sold. The source for the story is the disinterested
Miho museums curators who generously showed Pichikyan their sources,
described as nothing less than official documents.35 Translated from bazaar
archaeology, museum-speak, language, this signifies he was shown dealers
sales records. Simplistically, he allows that the story may appear too much
of a fairy tale for many people to believe in (his own wordsand a motif
borrowed from Barnetts defense of the OT-1 legend mentioned above), but
he believesand he thinks we should too: remember, he has seen the Millo
museums official documents.
Alas, the putative documents and the objects published and exhibited
document only that a lot of objects were sold to the Miho museum. That
an archaeologist can write such a nonsensical scenario and that a scholarly
journal would publish it says much more than we care to know about the
sad, solipsistic state of archaeological scholarship and publishing today.
Of the 2600 [or] over 2500 [or] approximately 3000 objects re-
corded in the Miho corpus, a good number are cataloged, some illustrated.36
These include five silver and gold statuettes of barsomcarrying priests,
other gold and silver statuettes, gold vessels, rhyta, and so forth, all dated
from the late 6th century bc to the Hellenistic period. This treasure in
Japan opens up new possibilities for studying the religion and art of the
Zoroastrians, and provides direct [from Japan] testimony demonstrating
that Zoroastrian temples had come into being as early as the Achaemenian
period.37 Interpretation of them seems to be fluid, for after noting that both
OT treasures are of identical character and could be interchangeable,
we are also told38 that their differences indicate they were manufactured in
different workshops: no archaeologist will disagree.
34 Pichikyan 1977, 309.

35 Pichikyan 1997, 309310.
36 Pichikyan 1997, 307, note 1, 311312.
37 Pichikyan 1997, 371, 378.
38 Pichikyan 1997, 380.
The full publication of OT-2 with complete color illustrations occurred

in 2002, in the Miho museums Bactrian catalogue. Pichikyan wrote only
a onepage essay here on ancient Bactria, and, together with B.A. Litvinsky,
another on Takht-i Sangin and OT-1 (213219), repeating the same belabored
claims already known, but nothing relating to OT-2. In essays written by oth-
ers, Pichikyans 1997 OT-1 and OT-2 terms are called the Bactrian Treasure.
The catalogue essays and the object description entries (presented anony-
mously) do not specify which objects comprise this Miho OT-2 treasure, but
we surmise that it consists of items number 43 to 210, as many were recorded
and described in 1997. Here a fifth gold priest (no. 48d) has been added with
no explanation (we await official documents, but I suggest they will show
that it was found recently by another grandson); also included here are illus-
trations of all the many gold decorated plaques, varieties of gold bracelets,
torques, earrings, pendants, gold and silver vessels, also coins, and so forth.
There are two brief articles in the catalogue by Anthony Green; whether
he wrote some or all of the catalogue entries is not revealed.39 Green tells
us what we already anticipate: all is well in Japan, the OT-2 is a genuine
hoard; all of its objects are ancient; most are of Achaemenian date; and
it is plausible to accept Pichikyans and B. Litvinskys campaign that
both the Miho and the British Museums Oxus Treasures derived from one
site, Takht-i Sangin, or at least the same geographical region. Green first
claims that OT-2 has certain remarkable affinities to OT-1, then records
many differences (as Pichikyan did earlier); further, he shares with us the
knowledge that both the crudity and the sophistication of the OT-2
material are not results of chronology, but of the circumstances of their
[ancient] creation, their mass production, a manufacture strategy to suit
all pockets, even that some were made by the donors themselves, or by
other amateurs. Green in addition provides his colleagues with an insight
into the mechanisms of Bazaar Archaeology Gender Studies, for he has
been able to recognize significant information about ancient Zoroastrian
gender customs revealed solely on OT-2 artifacts. Noting that some OT-2
gold plaques that depict figures bearing barsom are women (those in the OT-
1 corpus are men), he reveals that now, for the very first time ever, we know
beyond doubt (his ipse dixit command to us), that Zoroastrian women
carried barsom; barsom carrying was not a gendered activitybut he does
not produce a single excavated or textual example to prove his claim.40
39 Green 2002a, 220; idem 2002b, 22227.

40 Green 2002b, 226. While it is probable that female deities may have held barsorn (Yasht
5: 127), there is no evidence that human females did.
Greens article is basically a gloss on Pichikyans writings with a few more

museum-generated revelations. Both are equally worthless and embarrass-
ing to those who take archaeology seriously. The marvelthe Miho mod-
elwe are expected to celebrate here is not merely that additional OT mate-
rial might in the course of time surface, but that the event occurred more
than a century after the initial happening and at the very same site, that all
was kept together intact, and all brought across long distances and several
lands to England before arriving at its safe haven in Japan. Now welcomed
and curated here, it may be proclaimed to be one of the treasures of the
Miho museum, one of the treasures of the Japanese nation, one of the trea-
sures of the archaeological world. But while the Miho museum may claim
possession of an OT-2, it may not claim possession of an Oxus treasure.
IIa. Forgeries: OT-1
We turn to the chronology of the individual objects in the two museum col-
lections and ask: are all or some of the objects in the OT-1 collection ancient
artifacts, made by ancient craftsmen? The answer begins with what one
may truly refer to as an official document: not one object in the collection
was excavated, not one has a known provenience; each is a phenomenon
wi th no ancestry beyond provenance in bazaar stalls. Resulting from this
situation, we naturally ask of each object why is it genuine? and seek a
response by evaluation against what is known from manifestly genuine
that is excavatedartifacts. What results is that for a large number of the
objects in OT-1 (for OT-2, see below) no defensible evidence commands our
acceptance of them as ancient artifacts; in other words, they are most proba-
bly modern forgeries. The results from an archaeological and common sense
perspective are the same: if perchance they are in fact ancient we can not,
and may never, know.
As early as 1905 Dalton, along with his other reservations, raised the
issue of forgeries among the Oxus material. He claimed that the chalcedony
cylinder seal (no. 114) in the treasure had recently been copied in gold.41
The gold cylinder was seen and reported by Cunningham (it is not revealed
whether he owned it), but he did not mention the chalcedony seal.42 Dalton
41 Dalton 1964, xvi, no. 114.

42 Cunningham 1883, 258, pl. XXI, A; on page 259, pl. XXII, D, he published another, strange
but possibly genuine, Achaemenian stone cylinder seal but says nothing of its provenance:
it seems to be in the Bibliothque Nationale, Paris, for which see Ghirshman 1964a, 269, 432,
fig. 331.
dissimulates by avoiding any mention of Cunninghams role in publishing

as a genuine piece from the Oxus Treasure the gold seal, along with other
objects he had indicted as forgeries (see below). Bleibtreu rightly raised
the question as to whether Franks bought the gold seal from Cunningham
or another individual (dealer), a matter that speaks directly to the issue
of multiple sources, and the alleged unity and integrity of the treasure.43
Bleibtreu also published a gold plaque acquired by the Kunsthistorische
Museum, Vienna, in 1918, which has a scene that was obviously copied from
the chalcedony seal no. 114, and which she correctly indicts as a forgery; at
least two other scholars ask us to accept on their word alone that this forged
plaque is ancient Achaemenian.44 This plaque was of course made to be sold
specifically as another Oxus Treasure artifact found in the very same locus
as the OT material; it is a forgery just like those acquired by Cunningham.
Dalton, again not calling attention to Cunninghams role in purchasing
and publicizing them, also claimed that the silver decorated disc (no. 24),
and the silver caprid amphora handle (no. 10) from the treasure were also
copied in gold.45 He also recorded here that a gold, ancient, statuette (no. 2)
had been copied in Rawalpindi (no. 2a) and acquired by the British Museum
from the Franks bequest.46 Two more larger and clearly related statuettes,
with the very same form and clothing, but not, as nos. 2 and 2a, carrying
flowers, are known: one in gold was once in the dealer Brummers collection;
the other in silver is in the Vorderasiatisches Museum in Berlin.47
Curtis agrees with Daltons assessment and mentions still other forgeries
made in Rawalpindi: a gold plaque depicting a winged bull, gold bracelets
and a gold ring.48 But he does not address a methodological problem he
(and others before him), however inadvertently, raisedonly to ignore:
Cunningham purchased these forgeries from the very same dealers who
also sold him other objects that he, Dalton, Curtis, and others assume are
ancient and deriving from one locus on the Oxus River. Something is very
wrong here. If the forgeries said to come from the Oxus treasure site
by local bazaar dealers did not derive from the Oxus treasure site, how
then can anyone knowknow in the archaeological sensethat other
said to come from Oxus objects purchased from the same bazaar dealers
43 Bleibtreu 1998, 145, note 5.

44 Bleibtreu 1998, 148150, Abb. 2, 3; E. Haslauer in Seipel 1997, 119, 122, no. 172.
45 Dalton 1964, xvi, 2; Cunningham 1883, 260, pl. XXI G: whether this piece is in fact a
forgery, I cannot say.

46 Cunningham 1883, 65, pl. VI, A; Dalton 1964, pl. XIII.
47 The two are published in Pope 1958, 368, pl. 108 AC; no author.
48 Curtis 1977, 241.
and curated in the British Museum did derive from the same locus, and
belong to one Oxus Treasure? This conundrum has been evaded by those
defending the treasures integrityperhaps because confronting it leads to
contradicting the locus and chronology, the archaeological integrity of The
Oxus Treasure.
I add to the list another Oxus silver statuette curated in the Lands of the
Bible Museum, Jerusalem representing a sad-looking man holding a lotus in
his left hand, published as Achaemenian. In fact, it is a sad man and a very
bad modern forgery.49
Also discussed by Dalton was the matter of the authenticity of the gold
plaques.50 Again he accurately stated the problem: suspicion has been
aroused by the character of some of the figures, which are so rude and
grotesque that a child might have designed them a thought appropri-
ately echoed by Collon, but here employed by her as a defense of a forgery,
no. 63.51 And again Dalton turned away from such a view, because it pre-
cludes detailed criticisma phrase left unexplained, but in fact uninten-
tionally communicating that he was aware that no objective analysis of their
birth dates would ensue. A feeble rationalization is given instead: gold was
common in the area in ancient times and crudeness need not condemn
a piecethereby suggesting that crudeness guarantees their authenticity.
Curtis also allows that some of the plaques are extremely crude, but sin-
gles out only two as forgeries, nos. 63 and 64.52 R. Ghirshman picked up the
crude-defense theme to explain why he knew that the unique plaques of
the Oxus treasure and some plaques (which in fact are all forgeries) in
the Foroughi collection were genuine votive objects donated to a temple:
La qualit de lexcution est secondaire. Look to the metal to determine
ancient authenticity: gold = authenticity, an assessment also accepted by
Rehm.53
Inasmuch as the gold plaques have been declared genuine by fiat, they are
readily invoked by all Oxus Treasure believers as prime evidence to doc-
ument archaeological and cultural conclusions about votive gifts derived
from a putative temple. The most recent publication claims the gold
plaques seem to have been votive in character placed in a temple to rep-
resent the worshipper in the presence of the deity. The association with a
49 I. Ziffer in Seipel 1997, 119120, no. 166; other forgeries of Achaemenian art in the same
museum and catalogue include nos. 167, 192, 193.

50 Dalton 1964, 1920.
51 Collon 1995, 187.
52 Curtis 1997, 234; as noted, Collon 1995 did not challenge the former, 184, 187.
53 Ghirshman 1964a, 89; idem 1964b, 91; Rehm 1992, 10.
temple is reinforced by the fact that many of the figures are priests we
are dealing with a temple treasury 54 (italics mine). Earlier, Ghirshman
explained the differences in the garments worn by the figures represented
on the plaques as indicating different ranks; the plaques reveal class differ-
ences.55 The plaques thusmerely because they existreveal not only their
ancient birth dates and functions, they also inform us about social differen-
tiation in Oxus religious society.
Tendentious, after the fact, defenses of the plaques notwithstanding, to
the contrary, crudeness of style and execution as well as obvious misun-
derstandings of form and details are essential issues that most certainly do
not preclude critical review. Crudeness and apparent aberrations in ancient
(excavated) artifacts occur rarely, but we are here discussing born-in-the-
bazaar phenomena, non-excavated material, and quite a lot of it: and no
comparably crude Achaemenian material has to my knowledge ever been
excavated. The unparalleled nature of the plaques in the excavated cor-
pus is to be confronted solely on archaeological terms, without resort to
assumptions, fiats, beliefs and ad hoc cultural interpretations. What ensues
from such an investigation is that all essential details on the plaques
eyes, noses, hair, faces, hands, feet, hats, clothing, shoes, and sleevesare
completely unparalleled in any style and execution within any sphere of
ancient artmost certainly including Achaemenian, which seems to be the
intended cultural designation of their modern, neo-Achaemenian, creators.
Why are they ancient?
Plaque no. 48, first described as a priest by Cunningham, is glamorized by
Curtis as one particularly fine plaque , the finest of the gold plaques.56
The plaque is worked in low relief, apparently unique in the collection. The
plaque has other uniquei.e. unparalleledfeatures aside from a poorly
articulated Achaemenian style. Note the trousers, hat, skirt, mask, etc.; the
swords scabbard and hilt position are wrongly conceived and executed, and
the sword is not attached to the belt and thus floats in space; the mask shows
the chin and does not seem to cover the mouth; the garment either has short
sleeves, or the two lines at the wrists may have been meant to be the sleeve
ends, or braceletsbut in fact, this issue is but another indication that
the provincial craftsman had no idea what he was executing in the first
place. The figure also has two left feet, as does no. 49I have not found
two left feet on any figure in excavated Achaemenian art. Further, would a
54 Curtis 1997, 235.

55 Ghirshman 1964a, 94.
56 Cunningham 1881, 155, 158159, pl. XIV; Curtis 1997, 231, 241.
Zoroastrian priest be armedbeyond possibly carrying a knife to cut the

barsom? Cunningham compared the priests sword with that carried by
genuine Achaemenian soldiers, ignoring problems of style and also the mat-
ter that Zoroastrian priests do not carry both barsom and swords.57 In fact,
it can legitimately be proposed that it was a modern forger who copied an
Achaemenian soldier, adding barsom, by comparing excavated examples
of men bearing what may be barsom. These latter are never armedif in
fact they are carrying barsom and not clubs; and note among them what a
genuine face mask looks like.58 However, scholars have been expected to cite
plaque no. 48 as an OT-1 gift to the modern world that introduces hitherto
unknown information about ancient Zoroastrian priestly practices.
I raise one more matter. A number of the plaque figures carry what all
writers call barsom in their left or right hand. This ambidextrous custom did
exist in ancient representations (those just noted above), and the left hand is
also mentioned in ancient Zoroastrian texts. But here not a shred of ancient
workmanship is visible on any example. No. 38 is a problem for me, although
I tend to think it is not ancient: the figure if genuine surely represents
a female, not common in Achaemenian art, but the garments folds are
not typically placed, and the hand is quite large. From any objective, non-
museum-culture perspective, not one of the OT-l gold plaques can be readily
accepted as an ancient artifact.
A few other examples of Oxus art have appeared over the years, all are
published as ancient, and all are modern forgeries: Ghirshman reported on
three plaques from the Foroughi Collection, one gold, two of bronze, which
he said came from Luristan but ornes identique ceux des plaquettes du
Tresor de lOxus.59 A revealing comment. E. Rehm accepts Ghirshmans
revelation, publishing as ancient one of the three Foroughi plaques, adding
to the corpus yet another forgery of an Oxus barsom holder.60
IIb. Forgeries: OT-2
We now turn to the issue of the modern forgeries in OT-2, published in the
Miho museum Catalogue (2002). I shall not expend much time here, for we
have before us another museum-generated scenario. I do not claim that all
57 Cunningham 1881, 159.

58 Akurgal 1961, 173, fig. 120, 174, fig. 123; also Ghirshman 1964a, fig. 440.
59 Ghirshman 1964a, 90, pl. 18; idem 1964b, 91.
60 Rehm 1992, 10.
the many objects catalogued in the Miho museum under the rubric Bac-
trian Treasure or OT-2 are modern forgeries, but a large number clearly are.
Not unexpectedly, the most obvious of these are the many decorated OT-2
gold plaques, of which I argue not a single example can be accepted with-
out reservations to be ancient. All are crude and unparalleled in execution of
body forms and stylistic details; indeed many were spawned from OT-1 mod-
els, formally and as artifact models (bastards begetting bastards), especially
the barsom-bearing priests (Catalogue nos. 6466), although none here is
armed (nor are the OT-2 priest statuettes). The valuable new information
to be derived from bazaar archaeology is manifold.61 The OT-2 figures have
many varieties of clothing and trousers, some embellished with incised pat-
terns, varieties of hats that include a concave top, all have two left feet, all
have differently and strangely executed eyes, noses, hands, and shoes. The
other plaques depict a multitude of crudely executed figures, holding a vari-
ety of objects, and some have crowned heads. Especially insulting to those
viewing this material is the sun disc with a figure holding up his hands surely
in despair (no. 71), a fortress populated with females (no. 72), the chatting
scene (no. 81), and the agriculture scene (no. 86).
Not one of the original five, eventually six, gold and silver statuettes of
priests holding barsom, here indicated by thin strips of gold, can be
assumed to have had an ancient existence. They are fabricated as thick,
thin, tall, short and stocky, with thin arms, thick arms, straight and bow
legged, with fingered hands or blobs, with different hats, different sized
and executed face masks, and with different shaped and formed eyes and
eyeballsthe latter depicting dead, non-ancient men. Not one of these
lifeless creations could, would, have been made in an ancient workshop
the artisans would have been deported.
As for the human and animal statuettes (nos. 5162), who among us
can claim with certainty to know their birth dates, inasmuch as they are
unparalleled anywhere in the ancient excavated corpus? And how many of
the Achaemenian style bracelets, torques, and earrings (nos. 145186; surely
nos. 146, 170173) are modern forgeries or ancient is impossible to determine
without disinterested, non-museum generated laboratory testing: there are
scores of modern jewelers who could make any object on view in the Miho
museum. This cautious view obtains also for the plain gold vessels and ladle
(nos. 123144), all of which could easily have been produced in modern
61 For more information about ancient Zoroastrian religion revealed from forgeries see
Muscarella 2000, 82ff., notes 3739, and above.

goldsmith shops: the same hesitancy and caution obtains for the other
scores of trinkets from the OT-2 treasure. I have no idea what the many
blanks, body parts, and miniature bows and daggers signify, or whence
they derived. That there are a number of apparently genuine artifacts
plundered from now destroyed sitesthat have been collected, curated,
published and displayed under the OT-2 rubric is most probable.
Outside the walls of two museums in opposite corners of the world
there is no Oxus Treasure. Both OT-1 and OT-2 are modern constructions
documented by pastiches of stories and composed of ancient and modern
artifacts. Collectively. they neatly illuminate not the culture of the ancient
Oxus region but the cultural interests, ideologies and internal agendas of
modern artifact-acquiring museumsand these interests and agendas are
not those of archaeology, which seeks (should seek) to comprehend an
accurate context of material remains in order to begin to understand their
cultural history.
Bibliography
Barnett, R. 1968: The Art of Bactria and the Treasure of the Oxus. Iranica Antiqua
VIII, 3453.
Bernard, P. 1994: Le temple du Dieu Oxus Takht-Sangin en Bactriane: temple du
Feu ou pas? Studia Iranica, 81112.
Bleibtreu, E. 1998: Ein Chalzedon-Rollsiegel aus dem Oxus-Schatz und seine Nach-
ahmungen in Gold. Archologische Mitteilungen aus Iran und Turan 30, 145154.
Cellon, D. 1995: Ancient Near Eastern Art (Berkeley).
Cunningham, A. 1881: Relics from Ancient Persia in Gold, Silver, and Copper. Journal
of the Asiatic Society of Bengal 51, 151186.
. 1883: Relics from Ancient Persia, in Gold. Silver, and Copper. Second, Third
Notice. Journal of the Asiatic Society of Bengal 52, 6467, 258260.
Curtis. J. 1997: Franks and the Oxus Treasure. In M. Caygill. J. Cherry (eds.), A.W.
Franks, Nineteenth-Century Collecting and the British Museum (London), 231
249.
Dalton, O. 1964: The Treasure of the Oxus with other Examples of Early Oriental Metal-
Work (London).
Ghirshman, R. 1964a: The Arts of Ancient Iran (New York).
. 1964b: Le trsor de lOxus, les bronzes du Luristan et lart mde. In K. Bittel
et al., (eds.), Vorderasiatische Archologie (Berlin), 8894.
Green, A. 2002a: The Treasure of Bactria in the Miho Museum. In Miho Museum
Catalogue, 2002, 220.
. 2002b: The Gold Plaques of the Bactrian Treasure. In Miho Museum Cata-
logue, 2002, 225227.
Miho Museum Catalague 2002: Catalogue of Treasures of Ancient Bactria (Miho
Museum).
Muscarella, O. 1980: Excavated and Unexcavated Achaemenian Art. In D.

Schmandt-Besserat and O. Muscarella (eds.), Ancient Persia: the Art of an Empire
(Malibu), 2342.
. 1987: Median Art and Medizing Scholarship. Journal of Near Eastern Studies
46/2, 109127.
. 2000: The Lie Became Great (Amsterdam).
Pichikyan, I. 1992: Oxus-Schatz und Oxus-Tempel (Berlin).
. 1997: Rebirth of the Oxus Treasure: Second Part of the Oxus Treasure from
The Miho Museum Collection. Ancient Civilizations from Scythia to Siberia 4.
306383.
. 2002: Ancient Bactria: Mirage or Reality? In Miho Museum Catalogue 2002,
213.
Pope, A. (ed.) 1958: A Survey of Persian Art (London).
Rehm, E. 1992: Der Schmuck der Achmeniden (Munich).
. 2000: Votivbleche im 1. Jt.v.Chr. In R. Dittmann et alii (eds.), Variatio Delectat,
Iran und der Westen (Munich).
Seipel, W. (ed.) 1997: Land der Bibel. Katalog. Kunsthistorisches Museum (Vienna).
Zeymal, E. 1979: The Treasure of the Oxus (Leningrad).
chapter forty
EXCAVATED AND UNEXCAVATED ACHAEMENIAN ART*
Introduction
The bibliography of articles and books concerned with Achaemenian art

of the 6th4th centuries bc is large and grows continuously. Photographs
of selected reliefs from Persepolis occasionally illustrate their texts. But
more often than not there are far more illustrations and discussions of
the sumptuous arts of Persiagold and silver rhyta, bowls, bracelets with
animal headed terminals, whetstones, stone vessels, sculpture in metal,
stone or lapis, and so forthobjects that fill so many museums and private
collections. The mere publication or citation confirms them chronologically,
the provenience furnished in the text anchors them geographically. The
objects were manufactured in the Achaemenian period and were recently
excavated at known ancient sites.
But were they? Yes for some, no for others: for if one takes the time to
examine the pedigree, the acquisition history, of each of the many scores of
objects presented over the years as examples of Achaemenian art deriving
from named sites in Iran, Anatolia, Mesopotamia and other areas of the
ancient worlda responsibility accepted by few of those assuming the right
to publishan historical perspective other than the one formerly perceived
emerges. In the first place, one notes that there is a failure on the part of
many scholars to distinguish between an excavated object that is obtained
from an archaeological excavation, and an object known to the scholarly
community solely through the activities of an antiquity dealer or other
private individual. Both sets of objects are treated equally with regard to
derivation and archaeological value; each is indiscriminately furnished with
an ancient home, a site attribution, a provenience.
In the sense that it is possible to be objective in archaeological matters,
the excavated object by the nature of its derivation informs us of its prove-
nience; the unexcavated object by the nature of its derivation cannot. Yet,
* This chapter originally appeared as Excavated and Unexcavated Achaemenian Art, in
Ancient Persia: The Art of an Empire, ed. D. Schmandt-Besserat (Undena, 1988), 2342.
1044 chapter forty
because certain scholars are unaware of the dynamics of the antiquities

market and the manner in which much published material becomes avail-
able, and because they have been traditionally educated to accept at face
value museum labels and catalogues or dealer attributions, i.e. to treat them
as if they were site reports, fundamental questions are rarely asked: how
and when was the object acquired? Was it purchased or excavated? Where
and by whom? But questions unasked remain unanswered.1 And objects for
which there exists no proof whatsoever as to attributed sites are treated on
the same phenomenological terms and with the same historical implica-
tions as those for which there is objective evidence.
Furthermore, based on stylistic analysis and the study of details and tech-
niques, it may be concluded that a large number of the unprovenienced
objects in existence are genuine; they were manufactured in the Achaeme-
nian period and as such must be included in discussion of the art of the
Achaemenian Empire, albeit we no longer are in a position to establish
their ancient resting places. But, and perhaps more significant, this same
analysis also determines that a large number of the available unprove-
nienced material cannot be accepted as a product of any ancient artist, let
alone an Achaemenian; they are modern forgeries. Further compounding
the problem is the fact that many of these forgeries have been published
as found at specific sites, the names of which are generously supplied,
and which are themselves forgeriesforgeries of provenience. Doubly
damnedmanufactured by the forger to deceive, and gratuitously given
identification papers by scholarsthey function as a fifth column in the
archaeological realm.
Two questions generate themselves from this brief commentary. First,
does there not exist a quantity and quality of excavated Achaemenian art
sufficient enough to satisfy the needs of scholars attempting to study and
elucidate it? Second, is it possible to sort through the unexcavated but
genuine Achaemenian objects and arrive at viable conclusions, not based
on hearsay, with regard to their provenience? In part, as we shall see, the
answers to the questions will overlap, but in broad terms may be discussed
separately.
The first question is readily answered by consulting the list recently
compiled of excavated Achaemenian art (see Muscarella 1977a, Appendix
B). A large quantity of excavated fine Achaemenian art does exist: silver
1 Few authors, including the present writer, have avoided this error in the past. What
is argued here is that in the future traditional attitudes that accept received information as
necessarily reflecting reality be eliminated, and that rigorous attention to sources be the rule.
excavated and unexcavated achaemenian art 1045
and terracotta rhyta, silver, bronze, glass and stone bowls, gold, silver and
bronze bracelets and other types of jewelry, decorated sheet metal, textiles,
ivories, and various small items. Moreover, that list knowingly presented
as incomplete, should be expanded considerably, and to this issue we now
turn.
Appendix B (Muscarella 1977a) was compiled with the view to supplying
a corpus of those Achaemenian objects concerning which one could be cer-
tain came from specific sites.2 Maintaining a strict documentation of prove-
niences yields two facts, both objective and not subject to personal opinion:
it is not a question of believing, or judging, but of knowing (Ridgway 1977,
20). First, and foremost, all the objects derived from excavated sites are gen-
uine products of the Achaemenian period, and as such should serve as the
prime data that inform us about the nature and characteristics of Achaeme-
nian art. And it is this ordered category that forms the base from which
one analyzes and controls the unexcavated material. Second, any attempt
to record accurately the find spots and alleged occurrences of Achaemenian
art throughout the Near East and surrounding areas ought to beginand
endwith a strictly defined list of proveniences. Put another way, specific
objects from specific sites furnish the data for meaningful research. If the
data are not under control in a disinterested fashion, so to speak, if there is
no certainty that the objects are in fact what is claimed for them, the con-
clusions will be at the least misleading, at the worst, fictional.
I. Objects with Reasonably Secure Provenience
At the same time, I have come to the conclusion that certain important
objects deserve a separate listing almost equal to that in Appendix B. In this
proposed parallel listing the situation is such that while no unbiased author-
ity witnessed the actual discovery of the objects, all the evidence compels
us to conclude that the find occurred: if not necessarily from the suggested
site (allowing for caution), certainly from a recognizable area, e.g. some-
where in Bulgaria (at least the Balkans), or in the Caucasus, and so forth. It
is understood that some subjectivity is involvedhence the separate list
2 The claim made in Appendix B (p. 193) that it was compiled solely from material exca-
vated under controlled conditions was not intended to suggest that an archaeologist employ-
ing modern techniques supervised the excavation. Rather, the intent was to convey the
fact that some archaeologist or reliable authorityi.e. not merely alleged local peasants
witnessed the find. Thus, for example, although the silver rhyton from Arinberd was listed, it
was found during a construction operation and turned over to the authorities.
1046 chapter forty
but it can be kept to a minimum and under control if each find is analyzed
on its own merits based on the particular circumstances recorded in each
instance. Accuracy, not pedantry, is the goal (cf. Hamadan below). Thus, I
now believe that sumptuous Achaemenian finds credited to Duvanli in Bul-
garia, Prokhorova in the Urals, north of the Caspian, and Achalgori, and with
reservations, Kelermes in the Caucasus, may be accepted as Achaemenian
finds from these respective areas (cf. Muscarella 1977a, 192 f.).
The Duvanli find apparently occurred in 1925 and was in part dispersed.
The well-known silver amphora (fig. 1) was presented for sale in Sofia but the
handles were acquired separately in Duvanli itself, where a plundered tomb
was observed. Still more material surfaced both at Duvanli and in Sofia; and
in other instances objects acquired at one place matched those from the
other (Filow 1934, 40f.). On the basis of these facts it seems that we take
no risk in accepting the Achaemenian amphora and a silver bowl (No. 15,
fig. 60) as found where claimed; at the very least we may accept them as
from Bulgaria.
Rostovzeff (1922, 123) vigorously defended the finds from Prokhorova as
definitely deriving from that site. Indeed, the objects of particular concern to
us, two Achaemenian silver bowls and a seal, were secured by the Oranburg
museum. Surely it cannot be denied that the objects were found somewhere
in this remote area (cf. Altai), for if found further south, they probably would
have found their way to one of the market cities there, Tiflis, for example.
The Achalgori findsactually said to have occurred in the village of
Ssadseguri but sold to a dealer in Achalgoriwere tracked down within
a short time of their sale in Tiflis (Smirnov 1934, IX). To my mind, the fact
the E.S. Takaischwili did a creditable job of tracing the find, documented
by Smirnov, and because the objects were offered for sale in Tiflis, suggests
that they were found somewhere in the Caucasus, which is sufficient for our
needs. Included among the objects that concern us are gold earrings, gold
plaques, silver bowls and a silver amphora without handles, a unique object
(Smirnov 1934, pls. III, IV, VIIXII).3
The alleged finds from Kelermestwo superb gold bowls, one deco-
rated with repouss petals and lozenges, the other with repouss birds and
3 Of some interest is a silver gourd-shaped vessel with two holes on the same side,
48 f., No. 68, pl. XII, as it is exactly paralleled by five terracotta examples in northwest Iran.
Three from Hasanlu and one from neighboring Dinkha Tepe came from Iron Age II, late 9th
century bc contexts, and one was seen in a dealers shop in Rezaiyeh. Oscar White Muscarella,
The Iron Age at Dinkha Tepe, MMA 9 (1974), 78. Whether the silver example is Achaemenian
or earlier is not clear from the available information.
animals (Rostovzeff 1922, pl. VII; Schefold 1938, figs. 4, 5)present spe-
cial problems given the unfortunate circumstances both of the find and its
recording. Rostovzeff (1931, 278ff.) published succinct information regard-
ing the uncontrolled nature of the finds attributed to Kelermes: aside from
the controlled excavations of Wesselowski, one Schulz, a licensed digger
(an amateur, Rostovzeff 1922, 5), claimed in a report to the Archaeolog-
ical Commission of Russia to have made important discoveries. Rostovz-
eff describes Schulz notes as manchmal ungenauen, bisweilen absichtlich
erlogenen Berichten, and claims that he also melted down some of the gold
and sold it for personal gain. Schefold (1938, 9) also noted that the impor-
tant finds came nicht aus den beiden genau untersuchten Grbern, sondern
aus Raubgrabungen in der Nachbarschaft dieser Kurgane. Although accept-
ing the bowls as part of the Kelermes group he lists them as Unbestimmter
Herkunft.
From all this what can one conclude about the provenience of the bowls?
Given Schulz tendency to lie (if we follow Rostovzeff) we really do not know
whence they derived, Kelermes or elsewhere. At the same time, the cir-
cumstantial evidence surrounding the report suggests that the bowls came
from somewhere in the area of Schulz activities (could he have purchased
them and resold them to the authorities?). It seems to me that a caution-
ary attitude allows one to accept a Kelermes provenience only tentatively;
less tentatively, one may accept a general Caucasian provenience. Indeed,
the information available is far from decisive, and any use of the bowls as
Achaemenian objects from the Caucasus must be made with a caveat. (N.b.
that the bowl with the animal reliefs could be provincial Achaemenian or
perhaps even slightly earlier.)
The so-called Oxus Treasure, on the other hand, cannot by any stretch of
the imagination be accepted as a find from one specific site or even area;
to quote Dalton (1964, Preface), the story of its acquisition suggests more
readily the atmosphere of romance than the sober world of archaeology.
The story (stories!) attached to its discovery and recovery claims it occurred
over a period of years beginning in 1877 at more than one site, and that the
objects passed through many hands and traveled to many places, includ-
ing Afghanistan; some were even counterfeited. Thus each piece must ulti-
mately be treated as an individual object without a specific provenience,
rather than as a part of a single find (Dalton 1964 XIII ff., XVI; Barnett 1968,
34ff., Porada 1965, 168ff.).
Several other objects omitted in Appendix B have come to my attention
and should be noted. One is a bronze bracelet with animal head termi-
nals and the characteristic Achaemenian indentation opposite the opening,
1048 chapter forty
excavated at Khorsabad (Loud and Altman 1938, pl. 59, No. 127; Amandry
1963, 271, No. 14; Moorey 1971a, 220). Amandry, the source for other appar-
ent finds, suggested that another bracelet from Khorsabad (No. 123) is also
Achaemenian but this is not so certain. Amandry also refers to two gold
Achaemenian bracelets with lion-griffin terminals in the Izmir Museum
from the area of Magnesia (Amandry 1965, 585). The nature of the available
information suggests that the bracelets were acquired by the museum by
purchase or confiscation, rather than excavation. Be that as it may, if they are
Achaemenian as stated, we may assume that they came from somewhere in
western Turkeywith the reservation that assume is the governing word
here. Amandry (1963, 270f., No. 6) further reports a major find of Achaeme-
nian material that occurred in 1962 in Vani, Georgia, and which is now in
the Tiflis museum. To my knowledge, both the objects and the nature of
the find remain unpublished, but gold bracelets, earrings, a diadem, a pec-
toral (gorget?), and a bowl are mentioned. Until the objects are available
for examination we must reserve judgement, although given the widespread
dispersal of Achaemenian objects it would not be surprising if such a find of
Achaemenian material occurred in the Caucasus.4
II. Achaemenian Objects Documented

on Reliefs and Paintings
In addition to these portable, three dimensional objects qualifying for inclu-

sion on a provenience map, there are the reliefs from Persepolis. Here are
depicted in abundance and in detail a variety of examples of Achaeme-
nian sumptuous art. But depicted is a misleading term; illustrate is more
appropriate, as it best defines the archaeological value of the reliefs. That
the Achaemenian artists aimedand brilliantly succeededat creating in
stone an exact duplicate in all details of the model in hand has been hith-
erto recognized (Amandry 1958a, 17; 1959, 54; 1965, 583). Frankfort (1955,
4 Amandry (1963, 267, notes 2, 13) mentions a bronze ladle found in 1904 at Aydin, another
from Sardis, and a bracelet from Antioch, all in the Istanbul Museum. I have not been able to
get information on these objects up to the time this paper went to press (to meet a deadline).
Otto (1944, 10, n. 2) mentions two bracelets from Nippur, also in Istanbul. Ltfi Tugrul informs
me he can find no record of their existence. Kantor (1957, 21) cited as Achaemenian a stone
relief fragment of a striding lion excavated in Byblos. Half of the head is missing but enough
of the lion remains to allow one to conclude that the attribution is not certain. One would
expect belly hair wings and two raised areas under the eye; further the depiction of the
mane is not typically Achaemenian in execution.
232) pointed out that Achaemenian sculpture is more closely related to

the applied arts than that of most other countries and that the princi-
ples of Achaemenian sculpture are the same as these of the applied arts .
And Stucky (1976, 22) stated it neatly when he spoke of Die beinahe pho-
tographisch genaue Wiedergabe der Geschenke illustrated on the reliefs
(even if one disagrees with his view that the models were sometimes not
available).5 The objects illustrated represent three dimensional works, and
in this capacity serve as primary examples from an archaeological perspec-
tive of excavated amphorae (Delegations III, VI), vessels of various sorts
(Delegations I, IV, V, VII, VIII, XII, XIII, XV), bracelets (Delegations VI, XI),
swords and daggers (Delegations I, II, X, XVII), and fibulae (Delegation IX),6
and all excavated in Iran.
Outside of the two handled amphorae (fig. 2), the bracelets with lion-
griffin terminals (fig. 3), the swords and chapes, and the several vessel types
illustrated on the reliefs, no examples of these characteristic Achaemenian
objects are to date known from excavations within Iran. Nor, in the case of
some of the vessel types, are they known from excavationor clandestine
findsanywhere.
It may be countered that, although the evidence of the reliefs is indeed
fundamental for the documentation of proveniences, the wrong prove-
niences have been presented. Thus, inasmuch as the objects under review
are carried by subject peoples who brought the gifts from their provinces
to Persepolis, we cannot properly claim them as excavated in Iran, rather
they must be interpreted as deriving from the respective homeland of the
gift bearers; the provinces should be credited with the appropriate prove-
niences, not Persepolis. This position has been assumed by Amandry (1959,
54), Luschey (1938a, 761, 764, 770; 1939, 59f. No. 330), and Filow (1934, 201),
5 This fact explains why A. Godard, Le Trser de Ziwiye (Haarlem, 1950), 69ff., believed
that the same artists who carved the reliefs were also the gold workers; cf. Stucky 1976, 21f.
for the opinion that from the Achaemenian point of view two and three dimensional objects
were not considered to be separate entities.
6 For other fibulae represented at Persepolis and excavated there and at various sites in
Iran see Muscarella 1969, 283 (IX), 284 (XIX). For additional published references to fibulae
excavated in Iran see: Zendan AA 1965, 739, fig. 59 a, b; Baba Jan, Iran VIII (1970), 176; Bisitun,
AMI 3 (1970), 154 f., fig. 221, pl. 75:1; Ghalekuti, T. Sono, et al., Dailaman III (Tokyo, 1968), T5, T7;
Godin, T.C. Young, Jr., Excavations at Godin Tepe, First Progress Report (Toronto, 1969), pl. 123,
no. 2; Nush-i-Jan, Iran VII (1969), pl. X, b; Pasargadae, Iran I (1963), 41; War Kabud, Phoenix
XIV (1968), 123. fig. 42, left; Bastun, Iran VII (1970), 177; AMI 8 (1970), 51. pl. 33:1 (late). For the
Ziwye fibulae see Muscarella 1977b, passim. There are no fibulae from Sialk, cf. Moorey 1971,
174.
1050 chapter forty
who have maintained that Delegations III and VI are representations of

Armenians and Syrians respectively; consequently the objects they bear
indicate local manufacture and provenience in these areas. In fact, Amandry
(ignoring his sound comment that the delegations ne sont pas toutes iden-
tifies avec certitude, 1958, 17; see also 1965, 582 f.) has used his view that
Delegation III are Armenians to support an Armenian origin for several
unexcavated amphorae.7
Two arguments may be brought forth to refute this position. First, as a
matter of fact we do not know with certainty which specific peoples are
represented by Delegations III and VI. Elsewhere I have argued (Muscarella
1969, 283) that Delegation III could presumably represent either Armenians
or Cappadocians (categorically eliminating Delegation IX as the latter),
and that Delegation VI remains unidentified. But recently Schmidt (1970,
152f.) has presented good evidence to support the view that Delegation VI
represents Lydians, a position supported by Barnett, Porada, vanden Berghe,
and Hinz (see Muscarella AJA 75, 1971, 444). Thus if Delegation VI is Lydian
and not Syrian we would have a major geographical displacement with
regard to the alleged origin of the goods carried.
Second, in listing proveniences, it is obvious that we are solely concerned
with the final resting place of a given object, not with its putative place
of manufacture. Frankfort (1955, 230) perceptively stated the issue clearly:
Whether the different peoples bought or made such things [objects on the
reliefs] for the purpose of presentation to the king, or whether the Perse-
politan artists simply depicted the type of valuables with which they were
familiar, we cannot say. Further, Walser (1966, 25, 41, 72; see also Muscarella
1969, 282f.) has correctly pointed out that several Delegations carry similar
goods and that the reliefs are not necessarily meant to be in der Reihen-
folge eines administrativen Documentes, depicting only objects manufac-
tured by the particular people represented, rather they illustrate the bunten
Wechsel der Vlkervariett des Reiches (25). He further argues (72) that Die
weite geographische Streuung der Edelmetallgegenstnde zeigt, wie weit
die entwickelte Toreutik ber das Achmenidenreich verbreitet war. Wo
die einzelnen Ateliers lagen, msste die kunsthistorische Analyse genauer
7 Amandry in 1965, 583see also 1958b, 1140seems to have changed his mind and
accepted Delegation VI as Lydians. It is of interest that he did not compare the pair of
bracelets with lion-griffin terminals in the Izmir museum mentioned above with the same
bracelet-type carried by the Delegation VI tribute bearers. A pair of bracelets of the same
type is also part of the Oxus Treasure (Dalton 1964, pl. I, No. 116).
erbringen. The valuable objects carried by the Delegations could have come
from any part of the empire, not necessarily auf Ateliers in diesen Gebieten,
for these objects waren begehrtes Kaufgut des achmenidischen Reich-
sadels, das man sich wohl von weither kommen liess . With these opinions
I am in agreement; the final resting place of the objects illustrated on the
reliefs is our sole concern, because that is all we know, and that place is
Persepolis in Iran.
At the same time it should be noted that a number of scholars have not
neglected to discuss the illustrated works of art on the reliefs in context
with similar, three dimensional works: viz. Amandry in his definitive arti-
cles (1958a, 1959, 1963, 1965) on bracelets and amphorae; Luschey (1938a
and 1939) on bowls; Filow (1934) on amphorae; Moorey (1971a, 219, 313) and
Pudelko (19331934, 85) on bracelets; Kantor (1957) on bracteates; Otto (1944,
9 f., n. 2) on vessels and bracelets; Goldman (1951) and Stucky (1976) on
chapes; Walser (1966, Chapter V) for objects in general. But the literature is
also filled with missed opportunities. If the reliefs are considered at all it is
to cite them in passing as comparanda for three dimensional pieces, rather
than as the starting point of the discussion (see the chapters on Achaeme-
nian art in Ghirshman 1964, Culican 1965, Porada 1965).8 And there are puz-
zling examples where a scholar has corrupted his argument concerned with
an object or motif in Achaemenian art by citing unexcavated and spurious
objects rather than the reliefs where an appropriate parallel exists (Culican,
Chapter V, passim; Kantor 1977, 14, n. 16).
The horn rhyta9 and lion-griffin recognized on the reliefs of the Tomb of
Petosiris (Lefebvre 1923, pls. VII, VIII, IX), figs. 4, 5, a tomb of an official at
Hermopolis in Egypt, dated to ca. 310300bc, may not be quoted in the same
manner as that suggested for the Persepolis reliefs. For although Achaeme-
nian objects are clearly depicted they cannot be said to be illustrated. For
some reason no attempt was made to create an exact copy of a rhyton or lion
griffin with all details beinahe photographisch represented. Their value,
speaking only with regard to the issue under discussion, is not so much one
8 That the style of the relief documents is sometimes ignored may be illustrated by one
example. No one, to my knowledge, has pointed out that on the reliefs the animals on the
amphorae do not rest their feet directly on the vessel and that their heads face out; these
features are also noted on the Duvanli amphora and on the one depicted on the tomb wall at
Karaburun, Mellink 1973, pl. 44 (see also Muscarella 1977a, 179, no. 102105). Note that rhyta
are not represented on any Achaemenian relief, pace Ghirshman 1964, 252.
9 At least onepossibly morespecifically Achaemenian style rhyton is represented on
the reliefs, Adriani 1939, 357.

1052 chapter forty
of style but one of provenience; the reliefs inform us about the manufacture
of Achaemenian objects in Egypt around 300 bc (Muscarella 1977a, 194,
n. 100; Adriani 1939).
The rhyta and bowls depicted on the walls of the tomb have not been
ignored by certain scholars: Luschey 1939, 74, no. 422c; Adriani 1939, 352, 357,
359ff.; Svobada 1956, 12f., 61ff.; and Culican (1965, 153) specifically pointed
out that the tomb yields evidence for the manufacture of Achaemenian
objects in Egypt . But it is of some interest to note that Montet (1926),
who argued vigorously for evidence of Persian influence in the scenes on
the reliefs, as opposed to Greek influences suggested by Lefebvre, totally
ignored the rhyta and bowls; he stressed what he believed were Mithraic
elements in the bull sacrifice scene (Lefebvre 1923, Pl. XIX). Picard (1930)
correctly dismissed this opinion, but went out of his way to deny any Persian
influence, even for the characteristic lion-griffins on the bier. To Picard even
the rhyta could be Greek, and he made no mention of the Persian examples
available to him (cf. Adriani 1939, figs. 46).10
The same conclusions expressed for the Tomb of Petosiris objects must
equally obtain for the vessel with griffin headed handles depicted on the
wall painting at Karaburun in Lycia (Mellink 1973, pl. 44). Here too, judging
from the photograph, we have evidence of an Achaemenian amphora with
animal handles in an excavated context.11
To summarize the above conclusions, a map with excavated find spots
recorded must include: amphorae, swords and chapes, bracelets and vari-
eties of vessels at Persepolis in Iran; horn rhyta, vessels and lion-griffins in
Egypt; and an amphora in Lycia.
III. Conflicting Proveniences
We turn now to the second question posed above, concerned with recog-
nizing whether or not the proveniences offered for unexcavated objects are
verifiable or just convenient. The position argued here is that if an object is
10 Note that Adriani (1939, 359 ff.) also noted a Persian background for the Petosiris rhyta,
but he believed (360) they sono le stesse influenze che si riconoscona in altri prodotti di arte
egiziana contemporanea . He believed that the Persian examples are in fact to be dated to
the time of the Petosiris tomb, ca. 300bc, and were made in Egypt.
11 Because I was concentrating on objects that could be used stylistically both to define
Achaemenian art and to function as a guide for unexcavated objects. I left out the Petosiris
and Karaburun evidence from Appendix B.
in a private collection or was purchased by or donated to a museum, we can-

not legitimately speak about its provenience. If we are attentive to the aims
of archaeology it follows that these objects can play no role in discussions of
the geographical distribution of Achaemenian art. With objects in private
collections we can be certain that they have no known provenience; with
museum objects we cannot. And here, of course, we are back to our point of
departure: what value has the museum label or publication? Sometimes it
is possible to receive information about an objects acquisition history only
by communication with the museum authorities, although enqueries may
not always be answered.12 But it is a sad commentary on the state of the dis-
cipline if one has to resort to private communication to get significant and
basic information.
It is not possible, nor indeed necessary, in a short study to discuss the
pedigree of every Achaemenian object encountered in a United States or
European museum or private collection. It will suffice to examine the most
important sumptuous ones and to demonstrate that although not a single
example was excavated, each has been supplied with a geographical attri-
bution and adopted by scholarly consensus as an excavated object. These
attributed proveniences either have been accepted as received, or, in sev-
eral instances, have been shifted to another by a scholar claiming private
information from the archaeological underground, a type of samizdat not
vouchsafed to the rest of us.
The Rothschild silver gilt amphora was initially published as trouve en
Turquie, and, more specifically, trouve sur la cte de la Mer Noire, prob-
ablement [sic] entre Sinope et Trbizonde (Amandry 1958, 48, n. 84, 53,
pls. 24, 25). This attribution was accepted without question by Ghirshman
(1962, 79; 1964, 253) and Culican (1965, 125), but not by Tuchlet (1962, 85).
He challenged Sinope as the find spot, suggesting that a dealer may have
got it wrong; to him Cappadocia to the south seemed a more obvious area
for its provenience. Neither of these suggestions was satisfactory to Porada
(1965, 166, 242, no. 39) who moved the provenience further east in Anato-
lia, basing her claims on information supplied by a private reliable per-
son whose identity was not revealed.13 We have in hand an archaeological
12 My own experience has been more positive than negative. In writing this paper curators
from several museums have graciously and promptly answered my queries. Not to put too fine
a point on it, the answer given to my queries concerning objects said to come from X was
the one expected; see Muscarella 1977a, 160 ff.
13 Inasmuch as the amphora was not excavated by an archaeologist, the source of the
information could only be a dealer: whotelling different stories to different people

1054 chapter forty
anomaly, one object excavated in three separate areas: but, as we shall see,
not an uncommon phenomenon.
The pair of Graeco-Persian handles from a now lost amphora, one in
Berlin, one in the Louvre, have also been found, if we believe our sources,
in more than one place. For they proviennent selon certaines informa-
tions, dAmnisos (actuelle Samsoun), lest de Sinope et, selon d autres,
dArmnie (Amandry 1959, 53f.). Amandry himself had earlier tentatively
accepted the Armenian provenience (Amandry 1958b, 1141), and, indeed,
most scholars have elected Armenia as the place of origin (Luschey 1938a,
761, 764; 1939, 57, n. 330; Filow 1934, 203; Svoboda 1956, 47; Culican 1965, 125).
The ex-Borowski, Berlin amphora, on the other hand, has been allowed to
come only from somewhere in Iran (Amandry 1958b, 1141, fig. 8, recently
found in southern Persia; 1959, 43; Tuchelt 1962, 85); and the superb silver
rhyton-amphora in the Pomerance Collection was published as found in
Iraq, based on a dealers claim (Terrace 1966, 52).
Moving to other vessels, we find similar problems; in some instances an
example is accepted by consensus as deriving from one place, in others there
are plural proveniences. Thus, nearly all scholars have accepted as coming
from Syria the Berlin bronze rhyton (31158), a bronze amphora in a private
collection (Amandry?), and a cup with a bull protome purchased by Woolley
in Syria (Amandry 1958b, 1140, fig. 4; 1959, 44 ff.; n. 75; Culican 1965, 122, 125;
Moorey 1974, 159; Luschey 1938a, 762; 1939, 57, n. 330; Tuchelt 1962, 83, no. 1,
85; Zahn 1930, 148, fig. 1; Svoboda 46: with a slight reservation expressed for
this and other rhyton proveniences). The only one of these three vessels
that has a verified background in Syria is the one purchased by Woolley; but
where his Armenian vendor actually got it, and from whom (he claimed he
found it at Marash), must remain an open question.
All interested scholars have accepted the purchased Metropolitan
Museum silver horse-head rhyton (47.100.87, fig. 6) as coming from Mazan-
deran in Iran, probably on the dubious authority of Pope (1935, 1; Casson
1938a, 355; Otto 1944, 10, n. 2; Luschey 1938a, 763, n. 5; Tuchelt 1962, 59,
n. 13; Ghirshman 1964, 252). Nevertheless, all that is objectively known of
its history is that it belonged to the Jacks collection before the Metropolitan
Museum purchased it.
hardly qualifies as a reliable person. Compare the reliable people discussed by Ridgway
(1977), who prevented the truth about the Manios fibulas true proveniencea dealer
coming to light for almost a century.
Nor has Armenia been neglected in the distribution of sumptuous art. A

silver horn rhyton (with no hole) in the Louvre (3093) has been assumed
to come from there (Amandry 1958, 54; 1963, 270; Luschey 1939, 57, n. 330;
Tuchelt 1962, 83, n. 7, 85; Svoboda 1956, 46); Adriani (1939, 359) was the only
one to cite it correctly as di provenienza ignota. Svoboda (1956, 46) placed
the other Achaemenian rhyton in the Louvre (bronze, 3115) together with
the silver example as aus Armenien erworben, but Tuchelt (1962, 83 no. 2)
said it was In Baghdad erworben. In fact, all we know of this latter rhyton
is that it was bought from a dealer who had lived in Baghdad; of the former,
that is was bought from a dealer in 1897, who said it came from Erzerum
(note that the date coincides with the time that the Oxus Treasure surfaced,
some of which pieces, including the silver rhyton, were said to come from
Erzerum: Dalton 1964, XL).14
While these examples do not exhaust the number of stray objects casually
attributed to various sites and areas (one thinks also of silver bowls from
or said to come from Manzanderan, Syria and Sinope, silver objects from
India and Afghanistan, stone vessels from Persepolis, and so forth)15 the
point has been made. It is more important now to focus our attention on
two places where in the past many objects have been assigned, Hamadan
and Egypt.
IV. Objects Attributed to Hamadan
What objects may be isolated as factually deriving from Hamadan, as op-

posed to the many claimed to come from that site? The record yields infor-
mation on only two investigations conducted at Hamadan, both of which
have been incompletely published. In 1890 J. de Morgan stayed twenty-
four days at Hamadan and its surrounding villages (de Morgan 1896, 235,
259), during which time he recorded visible architectural features as well as
14 I wish to thank Pierre Amiet for sending me information on these objects and other
information.
15 For bowls see: Ackerman apud Casson 1938b, 371; Terrace 1962, no. 61; Amandry 1963,
260, 262, 271; Culican 1965, 120, 136: ILN, September 8, 1962, p. IV; Luschey 1938b, 78f.; 1969,
42, 49, 53; Silver: ILN, September 8, 1962, p. IV; Amandry 1959, 41, pl. 21: 4, 5; Stone vessels:
see Muscarella 1977a, n. 42a, p. 182, n. 83. In the Gazette Archeotogique 8 (1883), 237ff., pl. 41,
E. Babelon published a small stone relief from the Duc de Luynes collection, which was
acquis en Syrie o il a peut-tre t trouv . See also Herzfeld 1941, fig. 362; Amandry 1959,
42, n. 39, Achete en Syrie. The plaque is indeed Achaemenian and it is unique; its origin,
however, cannot be claimed with any certainty.
1056 chapter forty
various portable objects that he acquired. The small objects were illustrated
in seven plates, four (figs. 157160) were labelled objects trouvs dans les
ruines dEcbatane, and three (figs. 161163) as objects from Ecbatane et
environs. In no instance did de Morgan inform the reader how he acquired
the objects, that is, which, if any, had been personally excavated, and which
were purchased. The latter would have resulted from what de Morgan de-
scribed as une exploitation rgulire, for il n existe plus gure aujourd hui
de parties de lantique Ecbatane qui naient t exploites (de Morgan 1896,
236). But with regard to his method of acquisition nothing more was given
than the ambiguous statement about objects qu on recontre dans les ruines
de la ville (de Morgan 1896, 253). Therefore, although I do not think it too
strong to state that the objects published by de Morgan were acquired by
him at or near Hamadan, in the final analysis we do not know how many
were found, or acquired, in outlying villages, and how many were in fact
found within the city itself.
Calmeyer (1972, 65) accepts the de Morgan finds as bona fide material
from Hamadan, a position I find too secure given the inadequate informa-
tion available. From all the objects illustrated by de Morgan, and obviously
from different periods, Calmeyer singles out a bronze animal pendant, a
small gold figurine of a female, and a small bronze duck weight (figs. 160: 10,
157: 9, 159: 14) as Median or Achaemenian finds from the site. To my eyes, the
gold female is not Achaemenian, nor necessarily Median; nor is the pendant
easily characterized except as pre-Achaemenian; the duck weight, however,
is probably Achaemenian (cf. Schmidt 1957, pl. 82:4).
In 1914, a French expedition consisting of Fossey and Virolleaud con-
ducted excavations at Hamadan. Photographs of Le Bretons restored draw-
ing of fragments of a pre-Achaemenian spouted vessel (figs. 7, 8),16 and a
number of small animal figurines (Ghirshman 1964, figs. 122, 124; Calmeyer
1969, 102f., fig. 106; 1974, figs. 35), are the only objects made available for
study. While both groups of objects are clearly pre-Achaemenian, there is
no reason to assume that they are Median (cf. Calmeyer 1972, 65; 1974, 114).
These objects alone of all the many published as deriving from Hamadan
16 It will have been observed that the cast curved spout with the animal head terminal
of Calmeyer 1974, fig. 3 (Figure 8 here) is not the same spout in the restored drawing.
Pierre Amiet, who has generously allowed me to republish the Hamadan vessel fragments,
has informed me that Le Breton did not believe the spout belonged to the same vessel as
represented by the other fragments, and therefore drew the spout he thought appropriate.
If this is true, and Amiet supports the idea, we then have parts of two vessels excavated at
Hamadan.
may be accepted without reservation (cf. de Morgans finds) as deriving

from that site. Odd fragments of walls, bricks and column bases have also
been recorded by various travelers and should be noted (Dyson 1957, 31 ff.;
de Morgan 1896, 247; Luschey 1968, 115f.; Calmeyer 1972, 65; Schmidt 1953,
37f.).
A fairly large number of objects assigned over the years to Hamadan
remain to be considered: what is the evidence for this attribution? There can
be no doubt that for over a hundred years or more the local population of
Hamadan, joined by itinerant treasure hunters from other areas, have been
clandestinely exploiting the ruins by digging and probingand this is con-
firmed by all visitors to the site (aside from de Morgan, see Dyson 1957, 31;
Herzfeld 1930, 115). But short of a report from a reliable witness to an actual
recovering of an object,17 of which I know of no examples other than the
French expedition of 1914, it is impossible to know in a meaningful archae-
ological sense which, if any, of the many objects claimed by dealers, and
accepted or obliquely implied by scholars, to have come from Hamadan, in
fact did so. It is pertinent here to quote Moorey who was speaking to the
problem of Luristan and Amlash material, but whose comments apply to
all material on the art market: irresponsible and arbitrary attributions
by dealers more interested in pleasing customers than in preserving archae-
ological evidence, have all too easily lead to confusion (Moorey 1971b,
114f.; see also Muscarella 1977a, 158ff.; 1977b, 213 ff.). It is difficult to avoid
paraphrasing these comments: irresponsible and arbitrary attributions by
scholars more interested in pleasing colleagues or their own intuitions than
in preserving archaeological evidence .
In the AMI of 1930 (115ff.) Herzfeld claimed that a number of Achaeme-
nian objects were found at Hamadan: gold and silver tablets, a clay tablet,
a stone inscription, a silver vessel with an inscription of Xerxes, a cylinder
seal, and an inscribed bronze tablet. Interpreting from internal evidence
alone, because details are never given explicitly, it seems that Herzfeld did
not see these objects unearthed himself; he first saw them in the hands
of othersbut whether in Hamadan itself, he gives no clue. However, in
Herzfeld 1968 (238) he specifically stated that the inscribed bronze tablet
(fig. 9) was acquired by him in Hamadan (cf. Herzfeld 1938, 159, where he
merely stated that the tablet appeared in 1930), and that the owner claimed
it to have come from a place north of Hamadan, i.e. that it was brought
17 Even in such an instance there are problems: see Muscarella 1977a, n. 26; and note again
the problem with the de Morgan material discussed above.

1058 chapter forty
there to be sold.18 Of the other objects we cannot speak, but on Herzfelds

evidence it is known only that he saw and purchased the bronze tablet in
Hamadan, and that it may have come from a site to the north, the distance
of which was not given. Yet this purchased object becomes Die alteste sicher
einheimische Denkmaler of Hamadan (Calmeyer 1972, 65; cf. Calmeyer
1974, 113, where it is given as from Umgebung von Hamadan, which was
not what Herzfeld claimed).
A brief caveat is warranted here, at least parenthetically, concerning the
gold and silver tablets mentioned by Herzfeld (1930, 115; also 1926 and 1928)
and others over the years. I cannot find any reference in the literature indi-
cating that any tablet was unearthed in the presence of reliable witnesses,
nor one indicating that there is certain archaeological evidence that the
tablets came from Hamadan.19 That the tablets resulted from clandestine
digging is obvious, but from which particular Achaemenian site eludes us.
Put another way, if the tablets actually came from Hamadan we cannot
know in the sense that to know allows one to construct viable historical con-
clusions. Yet, no less an authority than Kent (1950, 107, 111, 113, with bibliog-
raphy) accepted them as found at Hamadan and assigned them H letters,
certifying them as excavated inscriptions from Hamadan (see also vanden
Berghe 1959, 109; Paper 1952, 169f.).
In the AMI of 1935 Herzfeld published four silver bowls with inscriptions,
all of which were dispersed to dealers and to the Kevorkian and Brum-
mer collections; some of the latter eventually came to the Ashmolean and
Metropolitan Museums (fig. 10) (Moorey 1974, 183 f., no. 181; Pope 1945, pl. 32;
Wilkinson 1949, 197,20 Mostafavi 1953, 95; ILN, April 16, 1955, 699, center
18 In Herzfeld 1941, 195, it was stated that the object was lost but it was purchased by
the Metropolitan Museum of Art in 1952 (52.119.12; fig. 9). How it got lost by Herzfeld and
acquired by a dealer is not known (by sale?). With regard to objects purchased in Hamadan
it is pertinent to recall a reference in Aurel Steins Old Routes of Western lran (London, 1940),
291, to antique seekers from Hamadan digging in Luristan. Surely they brought back their
booty to Hamadan to sell: which would then be listed by the purchasers in publications as
from Hamadan.
19 A reading of Herzfeld 1926, 1928 and 1930 produces no statement that Herzfeld himself
witnessed the find he discusses as follows: Beim Neubau einer Huschens kam ein antikes
fundament zutage, in dem die Goldplatte zwischen zwei durch ihre gute Bearbeitung aufflli-
gen Quadern endeckt wurde (1926, 2105), and In the town of Hamadan, there was discovered
a few years ago (1928, 1). Who was the source for this story? Where did Herzfeld encounter
it? Where did he first see the tablets?
20 On p. 187 of Wilkinson 1949 is the statement that in 1947 the Metropolitan Museum pur-
chased some of our finest pieces, but we had the opportunity of adding twenty-four antiqui-
ties from prehistoric and Achaemenian Persia, all from scientifically conducted excavations.
These last were acquired by exchange from the National Museum, Teheran . In the article
right). Here again Herzfeld neglected to reveal where he had viewed the
objects, but at the same time he volunteered no information regarding the
alleged find spot of the bowls (even in AMI 1937, where he was defend-
ing their authenticity). Nevertheless, in an anonymous article in the ILN,
April 16, 1955, 699, the Metropolitan Museum bowl (47.100.84, fig. 10, ex-
Brummer) as well as other objects acquired on the market by the museum
(below), were presented as gold and silver objects from Hamadan. Earlier,
Olmstead (1948, 353) also believed that the bowls were made at Hamadan.
But this attribution did not go unchallenged and he was cautiously ques-
tioned by Schmidt (1953, 37, n. 356): We must assume that Olmstead had
definite information with regard to the find-location of the bowls . Herz-
feld reported neither their find-spot nor their present location . Aman-
dry also challenged the Hamadan provenience for the bowls (1963, 271, n. 17)
but not for other objects.
By far the most important group of objects traditionally accepted as deriv-
ing from Hamadan is the collection of gold rhyta, bowls and daggers that
are now in the Metropolitan Museum and in Teheran. In the first publication
of the Teheran gold objects (ILN, July 21, 1956, 107) it was merely stated by
the anonymous author that The provenance of these objects is not exactly
known, but they are believed to have come from Hamadan; however, who
believed it, and why, is not revealed (cf. Mostafavi 1953, 145, 147; the bowl on
148 is listed by me as a probable forgery, Muscarella 1977a, no. 108). Of inter-
est to those who keep a record of shifting proveniences, is the gold bracelet
(ILN, July 21, 1956, 107 lower right corner) listed as reputedly from Hamadan
and conveniently dated to the sixth-fifth centuries bc; this bracelet and its
mate are usually assigned to Ziwiye and to the seventh century (as in ILN,
April 16, 1955, 699, center; see Muscarella 1977b, 200, n. 6).
Four gold Achaemenian objects acquired by purchase by the Metropoli-
tan Museum were categorically claimed to derive from Hamadan in the
anonymous ILN article of April 16, 1955 mentioned above: a gold bowl with a
Darius inscription (fig. 11), a buckle (fig. 12), a dagger (fig. 13), and a rhyton
(fig. 14) (54.3.14). In the same year Wilkinson (1955, 213, 220 ff.) published
these objects in more detail, describing them as from Hamadan, found
in the ruins of Hamadan, or with the modifications reputedly found in
three Achaemenian objects, one the silver Artaxerxes bowl, are published with no comment
regarding their acquisition. Lest there be doubt, none of these objects formed part of the
exchange; all three were purchased from the Brummer collection. The only Achaemenian
object given in the exchange was a bronze horsebit (48.98.19) of the same type as Schmidt
1957, pl. 78, no. 4, pl. 79, no. 7.
1060 chapter forty
Hamadan, and probably also from Hamadan. Transferring the specifica-

tions of provenience from the explicit to the less certain, especially after
initially claiming that all the objects came from Hamadan (213), might sug-
gest that perhaps the find spot of the objects is really not known. Unpalat-
able as this fact might be, it seems not to have prevented other scholars
from attributing the Metropolitan and the Teheran Museums gold objects
to Hamadan (Amandry 1958b, 1139f.; Culican 1965, Chapter VI; Ghirshman
1964, 252ff.; Porada 1965, 162, 164, 172, pls. 47, 42; vanden Berghe 1959, 109;
Calmeyer 1972, 66; Svoboda 1956, 49).21
In addition to these objects there is a large group of gold jewelry now
housed in the Oriental Institute in Chicago, in the Metropolitan Museum,
the British Museum, the Teheran Museum, and private collections, espe-
cially that of M. Vidal, that has almost universally been accepted as finds
from Hamadan (Muse Cernuschi 1948, 35ff.; Kantor 1957, 18 f., cf. 20 f.; Pope
1948, 58f.; ILN August 8, 1948, 215; Mostafavi 1953, 109 ff.; vanden Berghe 1959,
109, pl. 135; Barnett 1960, 29f. gives no provenience). Purchased objects all,
they have the same background as the material discussed above, which is
the antiquity market; they have no known provenience.
I do not think that at this point in our discussion it is necessary to
produce every stray object ever attributed to Hamadan; a surfeit of examples
can bore and exhaust the reader. Nevertheless, there are several important
objects that cannot be omitted and which deserve special attention. Their
interest to us is primarily with regard to the manner in which they were
published, for, as with the material already discussed, if we understand
how archaeological objects are published, how stating too much or too
little produces misleading information, we will begin to understand the
provenience problem.
Recently, Barnett (1968, 46, 47, n. 1) published from the British Museum
collection (BM 123270) an axe in the form of a lion, concerning which it
was stated that it was found at Hamadan in 1880 (see also Calmeyer 1969,
21 It is not clear why Calmeyer (p. 66) discusses a gold vessel in Cincinnati and a gold plate
in Teheran in an essay on Hamadan. Indeed, Calmeyer does note that the find spot of these
and other objects werden oft bezweifelt, but does not follow this thought to a conclusion
nor elucidate it; both pieces have been cited by me as probable forgeries (Muscarella 1977a,
nos. 109, 152). Cf. Luschey 1968, 117 who states of objects said to be from Hamadan: Jedoch
sind die Fundumstnde und die Herkunft nicht ganz gesichert. Moorey 1971a, 315 accepts a
Hamadan attribution for a silver bowl in the Ashmolean Museum as possibly correct in this
case, and suggests that it may be a rare example of a Median silverwork. It is now known
that the decoration (Achaemenian style) on the bowl is modern: Muscarella 1977a, no. 154.
41, Twith the wrong acquisition number). The facts about the acquisi-
tion are as follows, and I thank J.E. Curtis for sending me the following
information: The axehead was presented to the British Museum in 1880
by Sir A.W. Franks; it is said to have been found at Hamadan by a certain
C.P. Clarke. The juxtaposition of these two statements, one published, the
other from the Museum files, is instructive: an unverified claim from one
party to a second is transmogrified into an historical fact when published
by a third; Hamadan becomes Hamadan.
A stone Achaemenian head, probably from a capital (cf. Schmidt 1953,
fig. 55), was published as acquired by the Kansas City Museum from the
ancient Median and Achaemenid site of Ecbatana, (ILN, April 20, 1957,
642; Culican 1965, pl. 76). Vanden Berghe (1959, 110, pl. 137d) accepted this
claimand the implication that the head came directly from Hamadan to
Kansas City: Tout rcemment, on y [Hamadan] a exhum un fragment de
sculpture achemnide. To the casual reader the phrases acquired from
Hamadan and on y a exhum, might logically suggest that the tasks of
excavation and shipment were performed by an archaeologist rather than
a plunderer; they certainly imply that firm information was available to the
authorsinformation they could not have had, given that the piece was
sold by a dealer.22
Vanden Berghe further claimed that two bronze heads once in the Brum-
mer collection were found before the last war aux environs de Hamadan
(vanden Berghe 1959, 110, pl. 137a, b). The larger, and finer of the two heads
(fig. 15) was first published in the ILN, January 10, 1931, frontispiece, as an
Achaemenian King that was found near Hamadan, and is the source of
vanden Berghes attribution.23 Both heads were subsequently published by
22 Vanden Berghes volume could have been the most valuable work on Iranian archaeol-
ogy to date if only he had limited himself to excavated material. The unexcavated objects and
the hearsay evidence associated with them should have been placed in a separate section.
23 It will be obvious to anyone who examines the initial publication of objects in the
ILN that the magazine functioned as a laundry for the antiquities market. Inasmuch as the
magazinc has a well-deserved reputation as a publisher of material from recent excavations,
most scholars have assumed that everything published there has a legitimate pedigree, the
ILN seal of approval. ILN titles often suggest that actual excavation reports are to be discussed
when in fact plundered objects are being offered for sale: Recently Found Masterpieces
(December 17, 1958); Recently Found Treasures (July 17, 1948); Recent Discoveries (August
21, 1948); Recently Discovered Gold, Silver (April 2, 1960); Newly Found Masterworks
(May 23, 1959); New Discoveries in Iran (May 31, 1941), and so forth. And in some instances
the objects were apparently not recently found, but recently manufactured (see Muscarella
1977a, 160, n. 31, 163).
1062 chapter forty
Casson (1938a, 355f., pls. 105, 106, 107a, B) but with apparently two sepa-
rate proveniences, neither of them Hamadan. For the smaller head Casson
claimed in the text that It comes from Salmas near Lake Van but gave no
provenience for the larger. Puzzling, however, is the statement offered in the
previous footnote 3: This [the smaller head] and the other head are said to
have been found together, in Adharbayjan. These contradictory statements
may be understood, I suggest, by assuming that either Ackerman or Pope
added the footnote in his or her capacity as editor, for when Pope (1945,
17) published the heads he stated matter-of-factly that they were found in
the vicinity of Lake Urmiya, ignoring the earlier Hamadan and Lake Van
proveniences. Whatever the reasons for the two proveniences in his arti-
cle, Casson accepted both heads as products of the Achaemenian period
(not pre-Achaemenian as stated by Porada 1965, 233, n. 34), perhaps inad-
vertently echoing the original Hamadan reference.
The larger head was eventually purchased by the Metropolitan Museum
(47.100.80), where, following Pope, its provenience was given out confi-
dently as found in northwest Persia (Wilkinson 1949, 192 f.).24 This latter
attribution is the one presently accepted (Porada 1965, 62, 233, n. 34, fig. 38;
Calmeyer 1972, 65 also challenged the Hamadan provenience), as is the ear-
lier date suggested by Wilkinson. In fact the only certain provenience we
have for this mute masterpiece is that at one time it was in the Brummer
collection: nothing more or less.
V. Objects Attributed to Egypt
The problem of identifying with certainty which of the Achaemenian ob-

jects now or in the past housed in collections within Egypt and claimed
to have been found there, were actually discovered there, is more complex
than that concerned with Hamadan. It is usually assumed that if an object
was purchased in Egypt, it surely was found there, but this conclusion is
not always one subject to verification. Wakeling (1912, 122 f.), for example,
reported a Persian selling Near Eastern forgeries in Cairo; and Amandry
(1958, 20) put the matter succinctly when discussing a bracelet in the Louvre
that was acquired from a collection constituted in Egypt: ce qui ne prouve
pas quil y ait t trouv, ni, mme dans ce cas, fabriqu. The markets of
24 Its shifting provenience does not end here. One unpublished source gave the prove-
nience of the Metropolitan head as Tepe Tikhon, another as Chouchichi, near Lake Van.
Cairo have always attracted foreign merchants, especially given the number
of available westerners who seek antiquities. Therefore, when an Egyptian
museum catalogue cites an object with a provenience, but no documenta-
tion, or one with no provenience, it is difficult at this distance to establish
whether the object was excavated or purchased, or confiscated.
With these thoughts in mind, two bracelets in Egypt, and not mentioned
in Appendix B, should be considered as additional examples of Achaeme-
nian art in Egypt (unless local investigation proves otherwise). They were
published by Vernier (1927, 63, 188, pls. XVII, 52.148, XXI, 52.587; Amandry
1958, 14, n. 39, 20, n. 82) as deriving from Edfu and Mendes.25
Further, I believe it was an error on my part not to have cited as examples
of Achaemenian art from Egypt representations of jewelry on Egyptian
statuary, merely because their find spots are unknown. That the statues
were made in Egypt is beyond doubt. The first and best known example is
the torque on the statue of Ptah-hotep in the Brooklyn Museum (fig. 16),
(Cooney 1953a, figs. 1, 2, 5; Bothmer 1960, pl. 60, fig. 151; Amandry 1958a,
16, n. 55). This torque is a classic Achaemenian work and is paralleled by
a superb example in the Guennol collection (Cooney 1953a figs. 6, 7) and
by bracelets from the Oxus Treasure (Dalton 1964, nos. 136, 137). The second
example, less preserved, is a bracelet with animal head terminals on a statue
in the Vatican Museum (Botti and Romanelli 1951, 32 ff., no. 40, pls. XXVII,
XXVIII; Amandry 1958, 16, n. 55).26
But there are certain Achaemenian objects in western collections that
have been claimed to derive from Egypt and which were consciously omit-
ted from Appendix B. These include a limestone plaque in the Oriental
Institute depicting animal friezes (Frankfort 1950, 111 f., pl. III; Kantor 1957,
21, found in Egypt; Amandry 1958a, 16, n. 55; 1959, 42, n. 39); a limestone
plaque depicting a death scene, in the Hague (v. Bissing 1930, 226 ff., fig. 1a,
b; Svoboda 1956, 64); two stone lions said to have been found at Leontopolis
(fig. 17) (Cooney 1953b, 17ff., figs. 14; Amandry 1958, 16, n. 55), and a stone
25 The Toukh-el Garmous silver rhyton is obviously of Greek, Hellenistic, style and there-
fore cannot be included as an Achaemenian find in Egypt: Luschey 1938a, 770ff.; 1939, 74,
n. 422, 111, n. 632; Adriani 1939, 352, 359; Svoboda 1956, 64. Amandry 1958a, 16, n. 55; 1959, 51,
n. 101 refers to it as de type achemnide, which is true only in the sense that the rhyton is
based on Achaemenian types.
26 Both Botti (1956, 147 ff. and Amandry 1958a, 16, n. 55) believe that an Egyptian bust
in Florence wears an Achaemenian gorget. I know of no examples of gorgets represented

in Achaemenian art and therefore doubt the attribution. The torque illustrated on the
Alexander Mosaic is depicted to illustrate jewelry worn by the Persian king; it is not to be
considered an example of Achaemenian art recovered in the West.
1064 chapter forty
amulet (Cooney 1953b, fig. 7; Hoffmann 1958, fig. 6), in Brooklyn; a faience
rhyton fragment also in Brooklyn (Hoffmann 1958, fig. 3); another in the Lou-
vre (Roes, 1952, fig. 4; Hoffmann 1958, fig. 3); and finally a limestone lion head,
also in the Louvre (Roes 1952, fig. 7).27
What is the evidence that these objects came from Egypt? The Oriental
Institute plaque was said by its former owner to have been acquired in Egypt;
the Hague plaque nach unverdchtigen Zeugnis des Hndlers Cassira um
1909 in Mithrahine, der Statte der alten Memphis, zu tage gekommen; the
Louvre rhyton was once in the Clot Bey collection; as for the lion head, I
have not been able to get any data; unless there is solid information to the
contrary, it should be treated as an acquired, not an excavated, object. The
Brooklyn amulet, although listed by Cooney (1953b, fig. 7) in the photograph
caption as Egyptian, is presented in footnote 14 as Provenience unknown;
it was purchased from a dealer in New York City. The faience rhyton in
the same museum was originally given to the Peabody Museum in Salem,
Massachusetts, sometime after 1798 and subsequently, in 1948, was given as
a gift to the Brooklyn Museum (Hoffmann 1958, 10); in other words, it has
no known provenience. We are left with the lions in Brooklyn. According
to Cooney they were bought by Wilbour from a Dr. Fouquet of Cairo, who
had bought them in 1885, a year after they were said to have been found
by peasants and sold to a dealer. Collectively, the information we have
that the lions derived from Egypt is circumstantial. Nevertheless, certain
points are worth noting as they may help in determining provenience. The
lions were acquired in Egypt; they are large enough almost to preclude
the need to transport them from, say, Syria or points further east; and
they were said to have come from a sanctuary of the god Mahes, where it
would be appropriate to have representations of lions. It is therefore highly
probable that Brooklyns Leontopolis lions indeed came from that site and
they deserve at least a parenthetical listing among the Achaemenian objects
from Egypt.
At the same time, I still believe it is justified to have omitted the many
fine silver vessels said to have come from Pithom/Tell el Maskhuta (fig. 18);
I do not find it possible to connect firmly the silver with the alleged find
in 1947 (Cooney 1956, 43ff.). Here an Egyptologist might eventually be able
to clarify the nature of these objects and produce information that might
justify an Egyptian provenience.28
27 I would like to acknowledge the help of Ch. Desroches-Noblecourt and Robert Bianchi
for their cooperation in securing information on some of these objects.

28 I. Rabinowitz, Aramaic Inscriptions of the Fifth Century B.C.E. from a North-Arab
To summarize, it is suggested that in addition to the list of Achaemenian

finds from Egypt listed in Appendix B (Muscarella 1977a), we may add aside
from the objects on the Tomb of Petosiris above, two gold bracelets, the
torque and bracelet on the statues, and probably the Brooklyn lions.
VI. Achaemenian Objects from Greece
The number of Achaemenian objects excavated on Greek soil is surprisingly

small considering the extent of the Persian penetration amply documented
in the literature. Only four are known to me: a bronze helmet with a ded-
icatory inscription identifying it as Persian, and excavated at Olympia; a
horsebit from the Acropolis at Athens; a gold bracteate from Dodona and
another from Samothrace (Muscarella 1977c, n. 2 for references).29
Other Achaemenian objects have been accepted as found in the West but
without documentation or meaningful verification. An apparent Achaeme-
nian cylinder seal was claimed years ago to have been found (conve-
niently?), at Marathon, but, if true, the find cannot be verified, nor is the
seal presently available for study (Muscarella 1977c, n. 2). Claimed also for a
western provenience is a silver statuette in the Vorderasiatische Museum in
East Berlin said to have been found in Soli, Sicily (Casson 1938b, 352, pl. 108
A). Unfortunately, by the time this paper went to press I have not been able
to get information on this piece but venture to guess that it is a purchased
object.30 Luschey (1938b, 76, 78f., figs. 1, 2) claimed that a handle with two
Shrine in Egypt, JNES XV, I (1956), 19, accepts the objects in Brooklyn as definitely from
Tell el Maskhuta and reaches historical conclusions concerning the earliest known records
of Arab life in Egypt (p. 9). W.J. Dumbrell, The Tell el-Maskhuta Bowls and the Kingdom
of Qedar in the Persian Period. BASOR 203 (October 1971), 3344 comes to the same conclu-
sions. I wish to thank Richard Fazzini for these references.
29 The helmet and horsebit were not included in Appendix B because I was narrowly
concerned there with works of art, not artifacts in general.

30 While this paper was in press I received an answer to my question regarding the
provenience of the silver statuette from R.-B. Wartke of the Staatliche Museen zu Berlin,
DDR, whose cooperation I gratefully acknowledge. Wartke informs me that Die persische
Silberfigur, VA 4852 ist bei uns ohne Herkunftsangabe verzeichnet (Ankauf): nach Angabe
von Herrn Prof. Zahns aus Soloi/Cilicien. The italics are in the original; whatever the origin,
a western origin for the piece is excluded. Casson (or the editor) inadvertently wrote Sicily
rather than Cilicia.
Wartke also supplied me with the following information requested concerning a gold
bracelet aus Sardes published first by Luschey 1938a, fig. 5, and also by Otto 1944, notes 2
and 3; Pudelko 19331934, 8B; and Amandry 1958a, 14, n. 46: Das von Ihnen zitierte Stck
aus Sardes gehrt zu einem Paar Goldarmreifen und befindet sich jetzt in Antiken-Museum
1066 chapter forty
lion heads, which nach Angabe eines Vorbesitzers aus Thessalien stammt,
probably was actually made in Asia Minor, although it reached Thessaly
in antiquity. The heads look very much like pussy cats rather than lions
and to my eyes are not Achaemenian. Nevertheless, the use of unexcavated
objects to discuss ancient trade is unjustified on its own merits. I have
elsewhere discussed the gold bracelet in Karlsruhe as a purchased object
without provenience, although many have claimed it as an object excavated
at Corinth (Muscarella 1977a, 195).
Conclusions
Consolidating the evidence collected in Appendix B (Muscarella 1977a),

and the adjustments here, we record the following correct proveniences of
Achaemenian art. From the Black Sea area, including Russia and the Cauca-
sus: silver rhyta, vessels and bowls, perhaps gold bowls (Kelermes), sword,
some jewelry (the Vani finds to be added if necessary); from Greece and
the Balkans: silver rhyton, vessel with animal handles, amphora and bowls,
a gold bracteate, a bronze horse bit, a bronze helmet;31 Eastern Aegean
and Anatolia: silver rhyton, terracotta rhyta, vessel with animal handles
(Karaburun), silver bowls, glass bowl, gold and silver bracelets, jewelry; from
Mesopotamia: silver amphora handle, silver and bronze bowls, glass bowl,
bronze bracelet, earrings; from Syria-Palestine: silver and bronze bowls, sil-
ver bracelets, earrings, chapes, strainers; Egypt: gold/silver rhyta and bowls
(Petosiris), silver bowls, gold, silver, and bronze gilt bracelets, torque
(Ptah-hotep); Libya: glass bowl; Iran: from the reliefs: vessels with animal
handles, vessels, swords and chapes, fibulae; glass rhyton, silver bowls,
bronze bowl, glass bowl, carved stone vessels, stone, Egyptian Blue and
in Berlin (West)-Charlottenburg. Der Goldarmreifen war ursprnglich Besitz der Antiken

Sammlung der Staatlichen Museen, Inventar-Nr. 30989. Er ist zuletzt abgebildet und be-
schrieben bei A. Greifenhagen, Schmuckarbeiten in Edelmetall, Band II, S. 41, Tafel 35.
Herkunftsangaben: Geschenk von Th. Wiegand, 1904 erworben-angeblich von einem Bauern
aus der Nhe von Sardes. A second gold bracelet is published by Greifenhagen as having
been purchased at Smyrna (ibid., 41, pl. 35).
31 J. Harmatta, A Recently Discovered Old Persian Inscription, Acta Antiqua 2 (1954), 1
14, reports on an Old Persian Inscription on a clay tablet said to have been found in a garden
in 1937 in Gherla, Roumania. On pp. 13 f. Harmatta discussesbut rejectsthe possibility
that the tablet is a forgery (see below, Section III, no. 5); a thermoluminescence test of the
tablet would help resolve the issue. See also R. Frye, The Heritage of Persia (New York, 1963),
117, n. 96, who seems to accept the authenticity of the inscription.
bronze sculpture, gold bracelets, necklaces, earrings, bracteates, ivories, sil-

ver spoon, fibulae (see note 6).
The most common item in the repertory of excavated Achaemenian
art is the silver and bronze bowl, the great majority of which has been
recovered outside of Iran, and in every part of the Empire and beyond;
next in popularity is jewelry, in particular bracelets and earrings, also found
widely dispersed. Excavated amphorae with animal handles occur, except
for a handle in Iraq, only at Duvanli, and on the Persepolis reliefs and the
tomb painting at Karaburun; rhyta are more common, and are best known
from areas north of Iran and in Egypt (Petosiris). Textiles, gold and glass
bowls, animal headed cup vessels, and weapons (aside from the reliefs) are
rare. To date, ivories, carved stone vessels and sculpture are known only from
Iran.
Certain types of objects remain to be excavated: decorated mirrors (a
fragment of a plain example was excavated at Persepolis, Schmidt 1957,
103, pl. 81:53), gold rhyta, whetstones, ibex and other animal head pro-
tomes; elaborately carved stone vessels, metal bowls with a decorated cen-
tral medallion, and gold buckles and daggers. These types are known to us
at present only from the art market, where they deriveor in some cases
clearly do not derivefrom clandestine digging.
It will be obvious by this time that if we now compare the list of corrected
proveniences suggested in this paper to the larger one of putative prove-
niences assigned to unexcavated pieces, a different and a modified historical
picture emerges (cf. Amandry 1963, 270f.; Otto 1944, 9 f., n. 2). This is made
quite clear by concentrating on two characteristic Achaemenian objects, the
amphora with the animal handles and the rhyton, and adding or subtract-
ing them from the received provenience map. For the amphorae, Armenia
and Syria are removed; Anatolia and Iran legitimately remain because of the
Karaburun and Persepolis evidence. With regard to the rhyta, Armenia, Iraq
and Syria are removed; and Egypt (Petosiris) is added, with the understand-
ing that the evidence here is chronologically post-Achaemenian Empire.
As for the geographical distribution of remaining objects, the prove-
nience map remains basically the same, albeit the find spots are reduced in
numbers. Further, on the basis of the Persepolis reliefs, bracelets with lion-
griffin terminals, swords and chapes, and various bowl types are now added
to Iran.
If the aims of this paper have been achieved, it has been demonstrated how
facile it has been for uninvestigated and undocumented statements to rise
above the level justified by the evidence, thereby diverting scholars from
1068 chapter forty
the archaeological insistence for the specific and particular. Archaeology

cannot exist independently of what we make of it. The record is there to
prove that many of our past conclusions have been hostile to the data that
archaeological research has granted.
Diana Trilling has written of the need for negative space on the can-
vas of social and cultural events, a space wherein one might introduce,
without compromising accuracy, ideas and thoughts that morally and his-
torically circumscribe or enrich our understanding of the events depicted.
Archaeological research must become aware of its negative space, a space
that naturally surrounds each artifact and which contains the boundaries
and limitations for historical and anthropological conclusions, insights and
opinions, and even for leaps of intuition. While unexcavated objects pos-
sess some negative space which permits discussions of style, chronology
and beauty, only excavated objects contain the volume sufficient enough
to allow successful scholarly navigation, and to furnish an area in which the
full force of archaeological reasoning may maneuver. And it is solely within
the confines of this space that we are able to discuss provenience: where was
an object found, by whom, in what architectural or cultural context, juxta-
posed to what other objects; and only then may we speak about history.
Abbreviations
AA Archaeologische Anzeiger.
AfO Archiv fr Orientforschung. Graz.
AMI Archaeologische Mitteilungen aus Iran.
BASOR Bulletin of the American Schools of Oriental Research.
BICS Bulletin of the Institute of Classical Studies of the University of London.
BMQ British Museum Quarterly.
DLZ Deutsche Literaturzeitung.
ESA Eurasia Septentrionalis Antiqua.
ILN Illustrated London News.
JANES Journal of the Ancient Near East Society of Columbia University.
JAOS Journal of the American Oriental Society.
JFA Journal of Field Archaeology.
JNES Journal of Near Eastern Studies.
MMA The Metropolitan Museum of Art, New York.
RLA Reallexikon der Assyriologie.
ZFA Zeitschrift fr Assyriologie.
ZDMG Zeitschrift der Deutschen Morgenlndischen Gesellschaft. Leipzig.
Bibliography
Adriani, A.
1939 Rhyta, Bull. Socite Royale dArchologie Alexandrie X, 350362.
Amandry, Pierre
1958a Orfvrerie achmnide, Antike Kunst 1, 1, 923.
1958b Royal Persian Gold , ILN, December 17, 11391141.
1959 Toreutique achmnide, Antike Kunst 2, 2, 3856.
1963 Collection Hlne Stathatos III (Strasbourg), 260272.
1965 Orfvres grecs la Cour du Grand Roi? 8e Congrs Inter. dArchologie
Classique (Paris), 581587.
Barnett, R.D.
1960 Ancient Oriental Goldwork, BMQ XXII, 2930.
1968 The Art of the Bactrians and the Treasure of the Oxus, Iran. Antiqua VIII,
3453.
v. Bissing, F.W.
1930 Totenstele eines persischen Grossen aus Memphis, ZMDG 84, 226238.
v. Bothmer, Bernard
1960 Egyptian Sculpture of the Late Period (Brooklyn).
Botti, G.
1956 Buste di un Dignitarie della XXVII Dynastia nel Museo Egizio de Firenze,
Boll. dArte 2, 147149.
Botti, G., Romanelli, P.

1951 Le Sculture del Musee Gregoriano Egizio (Vatican).
Calmeyer, Peter
1969 Datierbare Bronzen aus Luristan und Kirmanshah (Berlin).
1972 Hamadan, RLA (Berlin), 6467.
1974 Zu einiger Vernachlssigten Aspeckten Medischer Kunst, Proc. of the IInd
Annual Symposium on Archaeological Research in Iran (Teheran), 112127.
Casson, Stanley
1938a Achaemenid Sculpture, Survey of Persian Art I, 346366.
1938b Achaemenid Metalwork, Survey of Persian Art I, 367376.
Muse Cernuschi
1948 Iran, Pices du Muse de Teheran (Paris).
1070 chapter forty
Cooney, John
1953a The Portrait of an Egyptian Collaborator, Bull. Brooklyn Museum XV, 2, 1
16.
1953b The Lions of Leontopolis, Bull. Brooklyn Museum XV, 2, 1, 1730.
1956 Five Years of Collecting Egyptian Art (New York).
Culican, William
1965 The Medes and Persians (New York).
Dalton, O.M.
1964 The Treasure of the Oxus (London).
Dyson, R.H., Jr.

1957 Iran, 1956, University Museum Bull. 21, 1, 2739.
Filow, B.D.
1934 Die Grabhgelnekropole bei Duvanli in Sudbulgarien (Sofia).
Frankfort, Henri
1950 A Persian Goldsmiths Trialpiece, JNES 9, 111112.
1955 The Art and Architecture of the Ancient Orient (Penguin).
Ghirshman, Roman
1962 Le Rhyton en Iran, Artibus Asiae XXV, 5780.
Goldman, B.
1951 Achaemenian Chapes, Ars Orientalis II, 4354.
Herzfeld, Ernest
1926 Eine neue Darius-Inschrift aus Hamadan, DLZ 42, 21052108.
1928 A New Inscription of Darius from Hamadan, Mem. of The Arch. Survey of
India 34, 17.
1930 Ariyarana Knig der Knige, AMI 2, 113127.
1935 Eine Silberschssel Artaxerxes I, AMI 7, 18.
1937 Die Silberschssel Artaxerxes I und die goldene Fundementurkunde des
Ariaramnes, AMI 8, 551.
1938 Bronzener Freibrief eines Knigs von Abdadana, AMI 9, 159177.
1941 Iran in the Ancient East (Oxford).
1968 The Persian Empire (Wiesbaden).
Hoffmann, Herbert
1958 Fragment of a Faience Rhyton, Bull. Brooklyn Museum XIX, 3, 1012.
Kantor, Helene
1957 Achaemenid Jewelry in the Oriental Institute, JNES XVI, 123.
Kent, R.G.
1950 Old Persian (New Haven).
Lefebvre, M.G.
1923 Le Tombeau de Petosiris III (Paris).
Loud, G., Altman, B.

1938 Korsabad II (Chicago).
Luschey, H.
1938a Achmenidisch-persische Toreutik, AA, 760772.
1938b Griechisch-persische Metallarbeiten in Antiquarium, Berliner Museen
LIX, 7680.
1939 Die Phiale (Bleicherode am Hertz).
1968 Lowe von Ekbatana, AMI (n.f.) 1, 115122.
Mellick, Machteld
1973 Excavations at Karatas-Semayk and Elmali, Lycia, 1972, AJA 73, 3, 293
303.
Moorey, P.R.S.
1971a A Catalogue of the Persian Bronzes in the Ashmolean Museum (Oxford).
1971b Towards a Chronology of the Luristan Bronzes, Iran, IX, 113129.
1974 Ancient Persian Bronzes (London).
De Morgan, J.
1896 Mission scientifique en Perse (Paris).
Mostafavi, M.T.
1953 Historical Monuments of Hamadan (in Persian) (Teheran).

1969 Review of Walser, Die Vlkerschaften auf den Reliefs von Persepolis in JNES
28, 4, 28285.
1977a Unexcavated Objects and Ancient Near Eastern Art, Bibliotheca Mesopota-
mica VII, Mountains and Lowlands, ed. L.D. Levine, T.C. Young, Jr. (Undena),
153207.
1977b Ziwiye and Ziwiye: The Forgery of a Provenience, JFA 4, 2, 197219.
1977c The Archaeological Evidence for Relations Between Greece and Iran in the
First Millennium B.C., JANES 9, 3157.
Olmstead, A.T.
1948 History of the Persian Empire (Chicago).
Otto, H.
1944 Ein achmedischer Goldwidder, ZfA 14 (48), 922.
1072 chapter forty
Paper, H.H.
1952 An Old Persian Text of Darius II (D2Ha), JAOS LXXII, 169170.
Pope, A.U.
1935 Recent Persian Art, ILN, March 2, p. 1.
1945 Masterpieces of Persian Art (New York).
1948 Recently Found Treasure , ILN, July 17, 5859.
Porada, Edith
1965 The Art of Ancient Iran (New York).
Pudelko, G.
19331934 Altpersische Armbnder, AfO 9, 8588.
Ridgway, D.
1977 Manios Faked? BICS 24, 1730.
Roes, Anna
1952 Achaemenid Influence upon Egyptian and Nomadic Art, Artibus Asiae, XV,
1730.
Rostovzeff, M.
1922 Iranians and Greeks in South Russia (Oxford).
1931 Skythien und der Bosporus (Berlin).
Schefold, K.
1938 Der Skythische Tierstil in Sdrussland, ESA XII, 278.
Schmidt, E.
1953 Persepolis I (Chicago).
1957 Persepolis II (Chicago).
1970 Persepolis III (Chicago).
Smirnov, Y.I.
1934 Der Schatz von Achalgori (Tiflis).
Stucky, R.
1976 Achmenidische Ortbnder, AA, 1323.
Svoboda, B.
1956 in B. Svoboda, D. Concev, Neue Denkmaler antiker Toreutik (Prague), 989.
Terrace, E.L.B.
1962 The Art of the Ancient Near East in Boston (Boston).
1966 Ancient Near Eastern Art, The Pomerance Collection of Ancient Art (Brook-
lyn Museum), 1357.
Tuchelt, K.
1962 Tiergefsse in Kopf-und Protomengestalt (Berlin).
Vanden Berghe, L.
1959 Archologie de lIran ancien (Leiden).
Vernier, E.
1907 La Bijouterie et la Joaillerie gyptiennes (Cairo).
1927 Catalogues Gnral des Antiquits Egyptiennes (Cairo).
Wakeling, T.G.
1912 Forged Egyptian Antiquities (London).
Walser, Gerold
1966 Die Vlkerschaften auf den Reliefs von Persepolis (Berlin).
Wilkinson, C.K.
1949 The Art of the Ancient Near East, Bull. MMA 7, 7, 186196.
1955 Assyrian and Persian Art, Bull. MMA 13, 8, 213224.
Zahn, Robert
1930 Neuerwerbungen im Antiquarium, Berliner Museen LI, 147149.
1074 chapter forty
Fig. 1. Silver Amphora, Duvanli, Bulgaria, Sofia.

Fig. 2. Persepolis Relief, Apadana, Delegation VI.

1076 chapter forty
Fig. 3. Persepolis Relief, Apadana, Delegation VI.

Fig. 4. Relief, Tomb of Petosiris, Hermopolis, Egypt.

1078 chapter forty
Fig. 5. Relief, Tomb of Petosiris, Hermopolis, Egypt.

Fig. 6. Silver Rhyton (courtesy Metropolitan

Museum of Art, New York, 47.100.87, Rogers Fund).
1080 chapter forty
Fig. 7. Bronze Fragments Excavated at

Hamadan (courtesy Muse du Louvre).
Fig. 8. Bronze Fragments Excavated at

Hamadan (courtesy Muse du Louvre).
Fig. 9. Bronze Tablet (courtesy Metropolitan

1082 chapter forty
Fig. 10. Silver Bowl (courtesy Metropolitan

Fig. 11. Gold Bowl (courtesy Metropolitan

Museum of Art, New York, 54.3.1, Dick Fund).
Fig. 12. Gold Buckle (courtesy Metropolitan

1084 chapter forty
Fig. 13. Gold Dagger (courtesy Metropolitan

Museum of Art, New York, 54.3, 4a, b, Dick Fund).
Fig. 14. Gold Rhyton (courtesy Metropolitan

Museum of Art, New York, 54.3.3, Fletcher Fund).
1086 chapter forty
Fig. 15. Bronze Head (courtesy Metropolitan

Museum of Art, New York, 47.100.80, Dick Fund).
Fig. 16. Statue of Ptah-hotep (courtesy Brooklyn Museum).

1088 chapter forty
Fig. 17. Lion, Leontopolis(?), Egypt (courtesy Brooklyn Museum).
Fig. 18. Silver Vessel, Tell el Maskhuta(?),

Egypt (courtesy Brooklyn Museum).

Archaeology Artifacts and Antiquities of The Ancient Near East PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Archaeology Artifacts and Antiquities of The Ancient Near East PDF

Încărcat de

Drepturi de autor:

Formate disponibile

Archaeology, Artifacts and

Antiquities of the Ancient Near East

The titles published in this series are listed at brill.com/chan

Muscarella, Oscar White.

Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

This book is printed on acid-free paper.

Section One. Iran

Section Two. Anatolia

Section One. The Aegean and the Ancient Near East

Section Two. Artifacts

Section Three. The Antiquities Market and the Plunder Culture

Section Four. Forgeries

Section Five. Forgeries of Archaeological Provenience

When asked by Jennifer Pavelko of Brill USA to compile a volume of my

My career as an archaeologist began with my interest in Egyptian archae-

SITES AND EXCAVATIONS

THE TUMULI AT S GIRDAN: A PRELIMINARY REPORT*

politan Museum of Art, New York. Reprinted by permission.

Hamlin while he was planning the site.

Figure 1. Tumuli I, E, and F, before excavation, right to left. Qalatgah lies to

Museum of Art, began a second campaign at Dinkha Tepe.5 Permission was

Figure 2. Plan of S Girdan showing Tumuli A to

Figure 3. Tumulus II, partially excavated, in the foreground;

Figure 4. Tumulus II, during excavation, and Tumulus

Figure 5. Plan of Tumulus I.

Figure 6. Section of test trench I, Tumulus I.

Figure 8. Circular cleavages in the surface of the main cut,

Figure 9. Cleavages in the east-west section

Figure 10. Plan of cleavages in the main cut, Tumulus I.

Valuable information is thereby supplied to us relating to the techniques

stones was encountered, this time, however, laid horizontally (Figure 7,

Figure 11. Stone revetment as seen in test trench 1, Tumulus I,

Figure 12. Plan of Tumulus II showing

Figure 13. Tomb of Tumulus II during excavation.

Figure 14. Tomb of Tumulus II after excavation.

Figure 15. Plan of the tomb, Tumulus II (Datum: +107).

Tumulus III was selected for excavation because it represents a medium-

Figure 17. Plan of Tumulus III.

Figure 18. East-west section of the southeast quadrant, Turnulus III.

Figure 19. Plan of Tumulus III showing excavated areas:

Figure 20. The excavated southeast quadrant of Tumulus III

Figure 21. Inner rock pile, Tumulus III.

Figure 23. Plan of the tomb, Tumulus III.

Figures 25, 26, 27. Stone whetstone-scepter,

3. A bronze knife blade with the remains of a plaited material adhering

Figure 28. Reconstructed drawing of a silver

The building of the tumulus may be reconstructed as follows: An area in

Tumuli whose construction exactly parallels that of the tumuli at S

(1938) p. 119, fig. 8a, b, p. 126.

are to be seen in dealers shops.

Figure 30. Sherds from the fill of Tumulus II.

Figure 31. Vessel with incised triangular decoration, Iran, VIIVI

the erection of the tumulus (the material was inadvertantly scooped up by

THE TUMULI AT S GIRDAN: SECOND REPORT*

politan Museum of Art, New York. Reprinted by permission.

Work was concentrated on the completion of the wedge-shaped trench

Figure 1. Rubble stones of Tumulus I. Note cleavage line at right.

Figure 2. East-west section of the northwest quadrant, Tumulus I.

Tumulus IV is the second largest in the S Girdan cemetery; it is about 7.5

Figure 3. Plan of Tumulus IV.

The stones were removed, and excavation continued to a depth of 2.5

Figure 5. East-west section of the southwest quadrant,

Figure 6. North-south section of the southwest quadrant, Tumulus IV.